QuantaStor Version ChangeLog: Difference between revisions

From OSNEXUS Online Documentation Site
Jump to navigation Jump to search
m Update security notices (qsn update)
 
(112 intermediate revisions by 3 users not shown)
Line 1: Line 1:
[[CATEGORY:support_guide]]
[[Category:Support guide]]
= Change Log Overview =
= Change Log Overview =
The change log contains a detailed summary of the changes made for each new release of QuantaStor.  Each change made to QuantaStor is tracked with an engineering ticket number in the form '''[QSTOR-NNNN]'''.  This information is provided so that if you're working with OSNexus support you can track the availability of a given product feature add, improvement, or maintenance fix to a specific release.  For information on how to upgrade your storage system please see the [[QuantaStor Upgrade Guide|Upgrade Guide]].
The change log contains a detailed summary of the changes made for each new release of QuantaStor.  Each change made to QuantaStor is tracked with an engineering ticket number in the form '''[QSTOR-NNNN]'''.  This information is provided so that if you're working with OSNexus support you can track the availability of a given product feature add, improvement, or maintenance fix to a specific release.  For information on how to upgrade your storage system please see the [[QuantaStor Upgrade Guide|Upgrade Guide]].
Line 27: Line 27:
* The latest version of QuantaStor can be found on our [https://downloads.osnexus.com downloads page here].
* The latest version of QuantaStor can be found on our [https://downloads.osnexus.com downloads page here].


* [[QuantaStor_Installation_via_USB_Thumb_Drive|Bootable Flash Drive instructions.]]
* [[QuantaStor Installation via USB Thumb Drive|Bootable Flash Drive instructions.]]


NOTE: Archived older versions of QuantaStor ISO install media and CLI tools can be found [https://downloads.osnexus.com.s3.us-east-1.amazonaws.com/list.html here].
NOTE: Archived older versions of QuantaStor ISO install media and CLI tools can be found [https://downloads.osnexus.com.s3.us-east-1.amazonaws.com/list.html here].
Line 33: Line 33:
= Recent Change Logs =
= Recent Change Logs =


== v6.3.2.TBD (April 15th 2024) ==
== New Security Notices ==
No security notices published since the latest release, v7.0.1.147 (November 17, 2026), affect QuantaStor. ''Last checked November 17, 2026.''


=== [[QuantaStor 6.3.2 Upgrade Instructions | Upgrade Instructions]] ===
== v7.0.1.147  (November 17th 2026) ==
Click [[QuantaStor 6.3.2 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.3.2 Upgrade Instructions | upgrading to QuantaStor v6.3.2]].
 
=== [[QuantaStor 7.0.1 Upgrade Instructions | Upgrade Instructions]] ===
Click [[QuantaStor 7.0.1 Upgrade Instructions | here]] for instructions on [[QuantaStor 7.0.1 Upgrade Instructions | upgrading to QuantaStor v7.0.1]].


=== Release Notes ===
=== Release Notes ===
Alerts and Configuration Analysis
* Fixed: Pacemaker node-down and node-up alerts not reaching QuantaStor on Ubuntu 24.04. [ QSTOR-12546 ]
* Fixed share quota Warning and Urgent alerts not being raised again when another share entered a free-space band that had already alerted. [ QSTOR-12827 ]
* Fixed: alert-config-set --pause-alerts accepting invalid pause periods and unknown alert types. [ QSTOR-12838 ]


Ceph Scale-out Block, Object and File
Ceph Scale-out Block, Object and File
* Added search capability to the S3 Bucket section of the Web UI. [ QSTOR-10288 ]
* Fixed S3 gateways that stopped answering requests while their service was still running not being detected or alerted. [ QSTOR-10646 ]
* Added Rule Sorting and an explicit rule remove button to the Storage Class Auto-Tiering Rules Dialog. [ QSTOR-10287 ]
* Fixed: a failed Ceph cluster create no longer leaves leftovers that block every later create on those nodes. [ QSTOR-12224 ]
* Added: You can now add the FQDN's for DNS names that you want the S3 Gateways to respond to via the Modify Realm/Zone Group Dialog. Note: you must add the FQDN's of the hosts or an FQDN alias e.g. s3.yourdomain.net (like used by DNS round robin) for the HAproxy load balancer to work properly. [ QSTOR-10285 ]
* Fixed: Object Pool Group creation failing when the Ceph cluster already had the default RGW zone group. [ QSTOR-12271 ]
* Fixed issues with Ceph S3 Bucket discovery with a large amount of buckets. [ QSTOR-10261 ]
* Fixed: S3 requests by hostname failing with NoSuchBucket on nodes with no DNS search domain. [ QSTOR-12749 ]
* Fixed some items related to Veeam SOSAPI integration based on testing in Veeam Certifcation scripts. [ QSTOR-10243 ]
* Fixed: cloud storage classes created in the web interface now write tiered objects to the destination bucket you choose. [ QSTOR-12857 ]
* Added: a Ceph-over-ZFS Optimized tuning profile. [ QSTOR-12869 ]
* Fixed: encrypted Ceph OSDs failing to start after a reboot that followed an OSD encryption key replace. [ QSTOR-12914 ]
* Fixed: the Ubuntu 22.04 to 24.04 platform upgrade on Ceph cluster nodes marking OSDs out while the node was upgraded. [ QSTOR-12917 ]
* Fixed a new Ceph RBD volume sometimes being mapped twice on the node that created it, which made deleting the volume fail with 'image has watchers'. [ QSTOR-12990 ]
* Fixed a force-deleted Ceph pool leaving its RBD volumes mapped on every node. [ QSTOR-12991 ]
* Fixed: the HAProxy load balancer not working on Ceph cluster members without a local RADOS gateway. [ QSTOR-13066 ]
* Fixed: encrypted Ceph OSDs failing authentication after a restart following a cephx key conversion. [ QSTOR-13126 ]
* Fixed: S3 auto-tiering for tenant buckets stopping after a coordinated Ceph upgrade from Reef to 20.2.x. [ QSTOR-13141 ]
* New Ceph clusters no longer enable discard on every OSD, and existing clusters have it turned off without restarting OSDs. [ QSTOR-13172 ]
* Delete Storage Class now opens on the right-clicked storage class, and STANDARD no longer offers Delete. [ QSTOR-13221 ]
 
Dashboards and Analytics
* Fixed: performance dashboards showing no data or wrong values for NFS and SMB shares and Ceph pools, OSDs, MDS and RGW. [ QSTOR-13128 ]
* Fixed: install-ceph-prometheus now runs on QuantaStor 7.0 (Ubuntu 24.04) and installs current Prometheus releases. [ QSTOR-13315 ]


Enclosures and Controllers
Enclosures and Controllers
* Fixed an issue where Dell RAID Controllers were being discovered twice in due to new support for the controllers discovery in Broadcom storcli utilities. [ QSTOR-10244 ]
* Fixed: a long-running disk identify blocking the request to turn the light off, and ignoring cancel. [ QSTOR-12785 ]
* Fixed: disk identify failing to blink drives on some enclosures. [ QSTOR-12844 ]
* Fixed: a failed disk identify blink now reports an error instead of Completed. [ QSTOR-12845 ]
* Added: an option to create Corvault disk groups and pools interleaved. [ QSTOR-12955 ]
* Fixed: enclosure auto-detection never selecting the Supermicro SC946 60-bay JBOD layouts. [ QSTOR-12966 ]
* Fixed: External System Profile Apply ignoring the Assign to Hosts setting. [ QSTOR-13068 ]
* Fixed: NVMe drives in the Supermicro ASG-2115S-NE332R getting no drive slot, so Identify could not light their bays. [ QSTOR-13103 ]


High Availability Failover
High Availability Failover
* Fixed an issue where rebooting an active High Availability storage node would cause ZFS encrypted shares to go offline for any active pools on that node prior to automatic failover. [ QSTOR-10218 ]
* Fixed a node placed in manual standby returning to Active when another cluster node's name starts with the same name. [ QSTOR-11917 ]
* Creating a site virtual interface right after creating its site cluster now succeeds, and a failed create no longer leaves the address active. [ QSTOR-12473 ]
* Unmanaged virtual interface messages and refused manual failovers now say when the site cluster is in maintenance mode. [ QSTOR-12475 ]
* Fixed a storage pool that had stopped writing to disk being reported healthy with no alert. [ QSTOR-12667 ]
* Fixed: when a pool export overran the HA group's Export Timeout, failover imported the pool without first fencing the exporting node. [ QSTOR-12671 ]
* Fixed: a storage pool import reporting Completed when the pool could not be imported. [ QSTOR-12674 ]
* Fixed: a service restart or one HA pool's failover moving another HA pool off its healthy node. [ QSTOR-12763 ]
* Fixed: stopping HA cluster services on a node giving up after 15 seconds and leaving the services in an inconsistent state. [ QSTOR-12834 ]
* Fixed package updates on Ubuntu 24.04 restarting HA cluster services, which could cause an unplanned HA pool failover. [ QSTOR-12913 ]
* Fixed site-vif-delete --convert-to-vif briefly dropping the IP address and raising HA group maintenance mode alerts. [ QSTOR-12931 ]
* HA nodes no longer report a failed lock manager service at every boot. [ QSTOR-13021 ]
* Fixed: adding a second cluster ring to an HA pair with clustered persistent reservations no longer hangs the peer node. [ QSTOR-13115 ]
* Fixed restarting Pacemaker on an HA node with clustered SCSI3-PR reservations taking minutes and leaving the cluster lock service failed. [ QSTOR-13175 ]
* Fixed: corosync and pacemaker restarting within a minute of an administrator's 'pcs cluster stop'. [ QSTOR-13227 ]
 
NAS Gateway / Cloud Containers
* Fixed: importing a cloud container onto an already-mounted export path leaving it without an NFS export or SMB share. [ QSTOR-12919 ]


Network Interfaces
Network Interfaces
* Fixed an issue where legacy VLAN interfaces that had the RX/TX optimizations flag set would not allow modification of the interface for any other setting. [ QSTOR-10272 ]
* Fixed: network ports no longer get renamed (eno1 to eno1np0) by kernel and distribution upgrades, which dropped IP, bond and HA VIF settings. [ QSTOR-11072 ]
* Creating a bonded network interface no longer leaves a second, conflicting definition for the same bond in the network configuration. [ QSTOR-12097 ]
* Fixed: an empty DNS search suffix breaking the host name and AD join, which now checks DNS identity and domain controller ports. [ QSTOR-13112 ]
 
Network Shares
* Kerberos NFS (sec=krb5) users on AD-joined systems now keep their AD identity instead of nobody. [ QSTOR-8198 ]
* Network share search is now case-insensitive, also matches share descriptions, and runs on Enter or as you type. [ QSTOR-10789 ]
* Added: AD join can use a pre-created computer account. [ QSTOR-11334 ]
* Fixed a share whose dataset mount hangs being retried every hour, leaking stuck processes and blocking other share operations. [ QSTOR-12664 ]
* Fixed a storage system removed from a network share namespace continuing to advertise the namespace over SMB and NFS until another namespace change. [ QSTOR-12921 ]
* Fixed: Changing a network share from the CLI no longer resets its owner group unless --share-owner-group is given. [ QSTOR-12945 ]
* Fixed share rename leaving a share disabled when an NFS client had a file open, and accepting conflicting names. [ QSTOR-12993 ]
* Fixed: AD join with a wrong administrator password now says the password is wrong. [ QSTOR-12999 ]
* Fixed copying share settings removing SMB user access from the source share, which could leave it open to all users. [ QSTOR-13176 ]
* Fixed: Network share owners and groups can again be set to a numeric UID or GID that has no matching user or group on the system. [ QSTOR-13320 ]
 
Quantastor CLI
* Fixed sys-modify silently ignoring --domain-suffix, and DNS server changes resetting a customized host FQDN in /etc/hosts. [ QSTOR-12116 ]
* Fixed: CLI options given as --option=false acting as if they were set to true. [ QSTOR-12782 ]
* Fixed: qs rg-create failing when --scope-type is left at its default. [ QSTOR-12916 ]
* Fixed: qs rg-res-add rejecting Object Tenants. [ QSTOR-12918 ]
 
Remote Replication, Snapshots and Backup Policies
* Replication schedules that reference a removed storage system link now switch to the current link instead of failing every run. [ QSTOR-11735 ]
* Fixed: a replication schedule rollback issued right after a replication run reporting Completed without restoring anything. [ QSTOR-12880 ]
* Fixed backup policies losing their destination sub-folder on modify and jobs not reporting failures or cancels correctly. [ QSTOR-12886 ]
 
SCSI / NVMe Target
* Added: Fibre Channel ports now report their negotiated and supported speeds. [ QSTOR-12855 ]
* Fixed: a Portal Group with no portals exposing its iSCSI volumes on every portal. [ QSTOR-12996 ]
* Added: an alert listing the iSCSI sessions closed when a Portal Group change removes a portal. [ QSTOR-13018 ]
* Fixed NVMe/TCP volumes created on a running pool not being offered to hosts until the pool restarted or failed over. [ QSTOR-13153 ]


Not Set
SNMP
* Added logic to prioritize high-speed data ports by default in the Ceph Cluster create and S3 gateway create dialogs. [ QSTOR-10266 ]
* Fixed: the SNMP agent not reporting new objects in a table that was empty when the agent started. [ QSTOR-12905 ]
* Fixed usable space calculation on Object Pool Group when there are multiple Pools or Storage Classes using the same OSD's. [ QSTOR-10264 ]
* Fixed an issue with modifying a network share after after a custom veto files setting had been added to the samba cifs options. Now the veto files merge as a set with the default .qssnaps veto files used by the VSS GMT snapshots feature. [ QSTOR-10249 ]


Security
Security
* Updated the OSNEXUS Security Notice service [https://services.osnexus.com/security/](https://services.osnexus.com/security/) for bionic and focal platform support. [ QSTOR-10269 ]
* Assigning Okta MFA to a user with no email is refused with a clear error. [ QSTOR-12885 ]
* Fixed: multi-admin approval requests not expiring on the node that created them. [ QSTOR-12981 ]
* Updated the gSOAP library to 2.8.146, the Genivia critical security release. [ QSTOR-13088 ]
* Added: an optional mitigation for CVE-2026-98432 (qs-util ipoptfilter enable). [ QSTOR-13148 ]
 
Server Integration
* Added: a QuantaStor storage plugin for Proxmox VE with snapshots, clones, migration, backups and multipath on iSCSI volumes. [ QSTOR-10624 ]
* Changed: the Proxmox VE plugin is now a native storage type installed from a .deb package and survives Proxmox upgrades. [ QSTOR-11575 ]
* Added LXC container support to the QuantaStor Proxmox VE plugin: container root disks can be placed on QuantaStor iSCSI block volumes. [ QSTOR-11968 ]
* Documented the guest I/O pause during a QuantaStor HA pool failover (about 30-45 seconds, no errors) and the iSCSI timeout guidance. [ QSTOR-12311 ]
* Documented that storage entries from the Proxmox plugin pre-release stop working on 1.0.0 and later. [ QSTOR-12487 ]
* Added API token authentication: a QuantaStor application token can be used in place of a password. [ QSTOR-12489 ]
* Fixed: Proxmox snapshot deletes failing with err=493 and leaving the guest locked. [ QSTOR-12537 ]
* Released Proxmox VE plugin 1.0.1. [ QSTOR-13004 ]
* Added: iSCSI multipath support in the Proxmox VE plugin. [ QSTOR-13005 ]
* Fixed: Proxmox container backups in snapshot mode failing on QuantaStor storage. [ QSTOR-13046 ]
* Added: full clones of a Proxmox VM from a snapshot, and fixed disk creation failing with "already exists". [ QSTOR-13047 ]


Service Core
Service Core
* Added support for replacing ntpd with chronyd in Ubuntu Focal platform based deployments. [ QSTOR-10239 ]
* Fixed: Send Support Logs reporting Completed when a slow collection or upload never delivered the bundle. [ QSTOR-12703 ]
* Fixed support bundles leaving out configuration changes made shortly before log collection. [ QSTOR-12705 ]
* Added the Recovery Manager: generate, download, upload and restore storage system configuration bundles from the web UI and CLI. [ QSTOR-13010 ]
* Fixed the NFS/SMB statistics collector logging sudo entries to the system log every 5 seconds. [ QSTOR-13011 ]


== v6.3.1.009 (April 1st 2024) ==
Storage Pool Management
* A storage pool is no longer marked offline when a disk-space query fails or times out while the pool is still imported and serving data. [ QSTOR-12804 ]
* Fixed: a serving storage pool being marked unmounted after a failed size check. [ QSTOR-13149 ]
 
Universal Containers
* A universal container service whose port is already in use now fails to start with a clear error, instead of repeatedly restarting Docker. [ QSTOR-12329 ]
* Fixed: a Service Container port conflict restarting Docker on the node. [ QSTOR-12898 ]
 
Upgrade Manager and Installer
* Added: SHA-256 and SHA-512 checksum files for each ISO, OVA and OVA zip download, alongside the existing MD5. [ QSTOR-11155 ]
* Changed: the Ubuntu 22.04 to 24.04 platform upgrade now requires QuantaStor 6.8.0 or later. [ QSTOR-12624 ]
* The Ceph upgrade now rejects an invalid target Ceph version with a clear error instead of ignoring it. [ QSTOR-12627 ]
* Fixed: Ceph upgrades failing or hanging on grids that reach the package repository through an HTTP/HTTPS proxy. [ QSTOR-12670 ]
* Added: qs-upgrade --cephPackagesOnly upgrades the Ceph packages on a system that is not part of a Ceph cluster. [ QSTOR-12775 ]
* Fixed: the coordinated Ceph upgrade on Ubuntu 24.04 restarting every Ceph daemon at once during the package stage. [ QSTOR-12851 ]
* Fixed: the coordinated Ceph upgrade reporting Completed when the post-upgrade cleanup failed. [ QSTOR-13146 ]
* Fixed: a coordinated Ceph upgrade refused with "Detected current upgrade in progress" after an earlier command-line upgrade. [ QSTOR-13163 ]
* Fixed: upgrades not refreshing the sudoers rules used by the NFS/SMB statistics collector and Pacemaker node alerts. [ QSTOR-13164 ]
 
Web Manager
* Long property values such as descriptions now wrap in the web UI property views instead of being cut off. [ QSTOR-9385 ]
* Fixed: ribbon bar buttons no longer draw borders in dark mode. [ QSTOR-9562 ]
* Fixed: adding a node to a running site cluster now completes and the node appears in the web UI. [ QSTOR-12005 ]
* Fixed: Modify Grid Time Zone Settings preselecting Europe/Andorra on UTC grids, and Etc/UTC can now be selected. [ QSTOR-12519 ]
* Fixed: tree searches now accept the same pool: and comma terms as the grids, and Physical Disks search finds disks. [ QSTOR-12819 ]
* Fixed % Utilized progress bars flashing white while grids refresh in dark mode. [ QSTOR-13222 ]
 
=== Security Notes ===
19 security notices published since v6.9.2.001 (November 2, 2026) affect this release: 13 fixed in this release, 2 fixed by a QuantaStor security update, 3 pending a fix, 1 pending a kernel update (13 externally-facing, 6 internally-facing). See [https://wiki.osnexus.com/security/?release=7.0.1.147 QuantaStor Security Notices for v7.0.1.147].


=== [[QuantaStor 6.3.1 Upgrade Instructions | Upgrade Instructions]] ===
== v6.9.2.001  (November 2nd 2026) ==
Click [[QuantaStor 6.3.1 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.3.1 Upgrade Instructions | upgrading to QuantaStor v6.3.1]].
 
=== [[QuantaStor 6.9.2 Upgrade Instructions | Upgrade Instructions]] ===
Click [[QuantaStor 6.9.2 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.9.2 Upgrade Instructions | upgrading to QuantaStor v6.9.2]].


=== Release Notes ===
=== Release Notes ===


Alerts and Configuration Analysis
Remote Replication, Snapshots and Backup Policies
* Fixed an issue where a pool running a scrub would report as degraded in the Config Analyzer health report. [ QSTOR-10258 ]
* Fixed: snapshot and replication schedules could stop running after upgrading to 6.9.1 on systems with many snapshots. [ QSTOR-13162 ]
 
=== Security Notes ===
No security notices published since v6.9.1.001 (October 28, 2026) affect this release.
 
== v6.9.1.001  (October 28th 2026) ==
 
=== [[QuantaStor 6.9.1 Upgrade Instructions | Upgrade Instructions]] ===
Click [[QuantaStor 6.9.1 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.9.1 Upgrade Instructions | upgrading to QuantaStor v6.9.1]].


Ceph Scale-out Block, Object and File
=== Release Notes ===
* Added qs-rgw-moveobjects script to allow easier s3api copy-object of a buckets contents to a different storage class. [ QSTOR-10268 ]
* Changed haproxy load-balancer to use balance source mode by default. [ QSTOR-10260 ]
* Fixed an issue with the Storage Class Auto-Tiering rules where multipart uploads with Capacity Operators of greater than or greater and equal to the Capacity Threshold would not be uploaded to the correct storage class. [ QSTOR-10254 ]


Enclosures and Controllers
Security
* Changed qs-util seagate-ap-split mode command to use zone 5 [ QSTOR-10259 ]
* Fixed: QuantaStor failed to start in FIPS mode after upgrading to 6.9 when a 3DES-encrypted license manager certificate was present. [ QSTOR-13041 ]
* Fixed: QuantaStor now distrusts the compromised eMudhra and Comodo AAA root CAs (USN-8944-1); please upgrade with security updates included (Core + Security, or qs-upgrade without --coreOnly) to install the change. [ QSTOR-13070 ]


Web Manager
=== Security Notes ===
* Increased default max bucket count for S3 user to 100K. [ QSTOR-10256 ]
10 security notices published since v6.9.0.464 (October 15, 2026) affect this release: 8 fixed in this release, 1 fixed by a QuantaStor security update, 1 pending a fix (5 externally-facing, 5 internally-facing). See [https://wiki.osnexus.com/security/?release=6.9.1.001 QuantaStor Security Notices for v6.9.1.001].
* Added new Grid Preferred Port status column in the Network Port Grid. It's hidden by default. [ QSTOR-10256 ]
* Fixed s3 gateway dialog height so it can show more nodes in the grid.  [ QSTOR-10256 ]
* Fixed s3 tenant create, if "Display Name" is not specified use the "Name" (display name is no longer mandatory in dialog). [ QSTOR-10256 ]


== v6.3.0.268 (March 20th 2024) ==
== v6.9.0.464 (October 15th 2026) ==


=== [[QuantaStor 6.3.0 Upgrade Instructions | Upgrade Instructions]] ===
=== [[QuantaStor 6.9.0 Upgrade Instructions | Upgrade Instructions]] ===
Click [[QuantaStor 6.3.0 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.3.0 Upgrade Instructions | upgrading to QuantaStor v6.3.0]].
Click [[QuantaStor 6.9.0 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.9.0 Upgrade Instructions | upgrading to QuantaStor v6.9.0]].


=== Release Notes ===
=== Release Notes ===


Alerts and Configuration Analysis
Alerts and Configuration Analysis
* Fixed an issue where md software raid os partitions where being reported in the config analyzer as offline storage pools. [ QSTOR-10168 ]
* Added: a Zabbix alert handler (Zabbix 7.0+) with an importable template of triggers and dashboards. [ QSTOR-10802 ]
* Fixed an issue where non-management ports were being flagged in the config analyzer for iSCSI acess being enabled. [ QSTOR-9891 ]
* Added: alerts can be paused. [ QSTOR-11665 ]
* Added: alerts for scale-out share capacity thresholds. [ QSTOR-11911 ]
* Fixed: report schedules did not save to a custom local save location. [ QSTOR-11937 ]
* Fixed: a removed disk could keep raising a Disk Health Non-optimal alert with a blank device and serial. [ QSTOR-12230 ]
* Changed: Discord is no longer listed by default in the Alert Manager ITSM integrations; it can be re-enabled in qs_alerthandlers.conf. [ QSTOR-12341 ]


Ceph Scale-out Block, Object and File
Ceph Scale-out Block, Object and File
* Blocked creating new OSDs when the OSD service is masked. [ QSTOR-10212 ]
* Added: replication of scale-out (CephFS) network shares to a second Ceph cluster using CephFS snapshot mirroring. [ QSTOR-9709 ]
* Show the S3 Object Storage Pool as the root node in the S3 User's and Tenant tree. [ QSTOR-10156 ]
* Added: bucket object lifecycle policies for external S3 endpoints. [ QSTOR-11603 ]
* Show the S3 Object Storage Pool as the root node in the bucket tree. [ QSTOR-10150 ]
* Fixed: client keyrings for volumes on erasure-coded pools now include the data pool, so direct RBD mapping works. [ QSTOR-11610 ]
* Add support for multiple data pools for storage classes for Ceph Object Storage Pool. [ QSTOR-10137 ]
* Fixed: RADOS gateway logs are now rotated daily and capped at 100 MB so they cannot fill the boot drive. [ QSTOR-11645 ]
* Add auto tiering capability to ceph buckets. [ QSTOR-10125 ]
* Added: per-bucket rate limits on operations and bandwidth for object storage. [ QSTOR-11836 ]
* Change the Ceph block storage pool to default to replica=3 for performance. [ QSTOR-10062 ]
* Fixed: deleting a scale-out object pool from Storage Management now points to qs object-pool-group-delete instead of failing with an unclear error. [ QSTOR-11896 ]
* For CephFS, the min_size for the metadata pool should be 2. [ QSTOR-10015 ]
* Fixed: retrying a failed Ceph node add no longer stops with "key does not match". [ QSTOR-11907 ]
* Introduce safe disk replacement for Ceph OSDs. [ QSTOR-9937 ]
* Fixed: qs cckr-assoc-get failed with a database column error. [ QSTOR-11924 ]
* Applied adjustments to Ceph Clusters for optimized scrubs on smaller and denser clusters. [ QSTOR-9900 ]
* Fixed: deleting a Ceph pool or CephFS share now removes it from client keyrings that referenced it. [ QSTOR-11925 ]
* Updated NFS Ganesha to 5.7 [ QSTOR-9821 ]
* Fixed: cloning a Ceph RBD volume or snapshot failed. [ QSTOR-11928 ]
* Add support for Ceph Multi-site bucket replication. [ QSTOR-6570 ]
* Fixed: a failed Ceph RBD clone can now be retried, and force-deleting a Ceph pool removes its volume records. [ QSTOR-11929 ]
* Fixed: qs osc-create rejected some recommended storage class names, such as EXPRESS_ONEZONE and OUTPOSTS. [ QSTOR-11943 ]
* Fixed: load-balanced object storage now redirects ports 80 and 443 to the load balancer's ports instead of falling back to a single gateway. [ QSTOR-12073 ]
* Fixed: qs object-user-create --tenant failed with "Unable to locate tenant". [ QSTOR-12123 ]
* Changed: bucket and S3 user discovery now uses the RGW admin API, so large namespaces scan much faster and a failed gateway reply no longer removes buckets or users. [ QSTOR-12133 ]
* Fixed: qs bucket-get now finds tenant-scoped buckets by bare name or tenant/bucket. [ QSTOR-12148 ]
* Changed: Ceph updated to Tentacle 20.2.4 on Ubuntu 22.04 and 24.04. [ QSTOR-12162 ]
* Fixed: a deleted object storage class data pool was recreated by the gateway; deleting the default STANDARD pool is now refused. [ QSTOR-12165 ]
* Added: Ceph over ZFS - zvols in a ZFS storage pool (including dRAID) can be used as Ceph OSD devices. [ QSTOR-12166 ]
* Added: storage classes can be deleted. [ QSTOR-12178 ]
* Fixed: upgrading RGW multisite deployments to Ceph 20.2.4 keeps zones replicating, and secure SigV4 is restored once every zone is upgraded (CVE-2026-54330). [ QSTOR-12269 ]
* Changed: Ceph 20.2.4 clusters keep legacy CephX keys so kernel clients keep working; qs ceph-cluster-key-convert converts daemon keys to aes256k (CVE-2025-30156). [ QSTOR-12270 ]
* Fixed: RGW admin operations now use HTTPS when the gateway runs in HTTPS mode. [ QSTOR-12294 ]
* Changed: Create Ceph Cluster reports progress per stage, and the Create OSDs dialog keeps selections when the search filter changes. [ QSTOR-12323 ]
* Fixed: Ceph OSD creation now takes seconds instead of minutes on large devices. [ QSTOR-12332 ]
* Changed: the default metadata DB size for new Ceph OSDs is sized from the selected devices, up to 3% of raw capacity. [ QSTOR-12461 ]
* Added: cephfs-mirror and rbd-mirror are installed on Ubuntu 22.04 and 24.04 appliances. [ QSTOR-12547 ]
* Fixed: Ceph OSDs on a loopback-alias public address showed Offline, and a missing OSD now raises an alert. [ QSTOR-12734 ]
* Fixed: qs-s3util move-objects now rejects a storage class that does not exist (override with --skip-class-check). [ QSTOR-12755 ]
* Fixed Replace OSD losing encryption and DB/WAL devices; the replacement disk is now cleared automatically and disks in use are refused. [ QSTOR-12894 ]
* Fixed encrypted Ceph OSD key replace failing on grids with journal devices, CLI OSD create dropping DB/WAL devices when no size was given, and failed OSD creates leaving orphaned OSD IDs. [ QSTOR-12901 ]
 
Dashboards and Analytics
* Added: External Dashboards - custom dashboards embedded in the Web Manager and organized into groups. [ QSTOR-9308 ]
* Added: a View Report option to view pool reports in the web interface. [ QSTOR-9707 ]
* Added: a network share usage tracker and dashboard, filterable by user or group. [ QSTOR-11720 ]
* Added: a Replication view on the Share and Volume dashboards that charts replication transfer rate. [ QSTOR-12289 ]
* Fixed: dashboard tile actions are now keyboard accessible and their colours meet WCAG AA contrast. [ QSTOR-12491 ]
* Changed: Trends charts draw lines instead of filled areas, legends toggle series again, and the pool chart plots used space against capacity. [ QSTOR-12575 ]
* Fixed: volume capacity and trend statistics were not recorded when optimized stats collection was enabled. [ QSTOR-12576 ]
* Fixed: Trends and Usage windows of 12 hours or longer showed zero capacity. [ QSTOR-12577 ]
* Fixed: updated statistics rollup settings were not applied to upgraded systems. [ QSTOR-12578 ]


Disk Management
Disk Management
* Added support for ScaleFlux CSD 3000 SSDs that support builtin compression provisioning. [ QSTOR-10198 ]
* Changed: removed unnecessary SMART query log messages on CORVAULT systems. [ QSTOR-11583 ]
* Fixed an issue with disk correlation by serial number on Broadcom/LSI 93XX series HBAs. [ QSTOR-10194 ]
* Added: a [tag:VALUE] disk filter, so tagged disks can be selected by tag, for example in pool-create --disk-list. [ QSTOR-11887 ]
* Show osd number in 'qs disk-list'. [ QSTOR-9717 ]
* Added: a Mode option for the disk performance test to run disks serially, in parallel, or grouped by vdev or enclosure. [ QSTOR-11953 ]
* Fixed: healthy SATA drives behind a SAS HBA raised false non-optimal health alerts. [ QSTOR-12082 ]
* Fixed: drives behind an HPE Smart Array Gen10 HBA (smartpqi) were not shown on Ubuntu 24.04. [ QSTOR-12726 ]


Enclosures and Controllers
Enclosures and Controllers
* Add enclosure view for the ASUS 2u24 NVMe. [ QSTOR-10202 ]
* Fixed: a Seagate CORVAULT that was not fully populated was given the wrong enclosure layout, or none. [ QSTOR-12088 ]
* Fix controller selection issue in the 'Identify Hardware Controller Disk Device' dialog for when the selected object in the Physical Disk tree is a Physical Disk and not the Disk itself. [ QSTOR-10195 ]
* Added support for the Dell PowerEdge R740xd 24-bay NVMe server, including its enclosure layout and drive slot mapping in Hardware Enclosures & Controllers. [ QSTOR-12944 ]
* Updated wddcs tool to 3.2.8.0 [ QSTOR-10166 ]
* Added to feature to create or delete multiple External System Volume configurations. [ QSTOR-10061 ]


High Availability Failover
High Availability Failover
* Fixed an issue with the Create HA Failover Group Dialog when there are active FC sessions on volumes on the selected pool. [ QSTOR-10169 ]
* Validated: Windows Server 2022 Hyper-V failover clusters with Cluster Shared Volumes on QuantaStor Fibre Channel HA pools. [ QSTOR-10677 ]
* Changed NVMe reservations to use cptpl=2 when available so that iofencing is cleared on power on of a nvme drive. [ QSTOR-10124 ]
* Added: clustered SCSI-3 persistent reservations for HA storage pools, so client reservations (Hyper-V, Windows Server Failover Clustering) survive failover. [ QSTOR-10705 ]
* Fixed: service SPNs on an HA VIF's AD computer object now keep working after failover. [ QSTOR-10884 ]
* Fixed: moving a site cluster VIF to a node with no location constraint now fails at once with an explanation instead of after a minute of retries. [ QSTOR-11605 ]
* Fixed: deleting an HA VIF from an active pool could fail the pool over to the standby node. [ QSTOR-11694 ]
* Fixed: Set Grid Sync Master reverted within seconds on grids with a Grid Management VIF. [ QSTOR-11698 ]
* Fixed: the FC link-state failover policy now applies only to pools that export volumes over FC. [ QSTOR-11855 ]
* Fixed: an HA node lost its manual-standby state after a reboot or distribution upgrade. [ QSTOR-11957 ]
* Fixed: deleting an HA pool or HA VIF could trigger a spurious failover on the peer. [ QSTOR-12002 ]
* Fixed: a stale site-cluster association could produce a corosync configuration that stopped corosync from starting. [ QSTOR-12026 ]
* Fixed: a manual-standby setting left over from an earlier configuration could force HA nodes into standby after an upgrade. [ QSTOR-12044 ]
* Fixed: failed resource actions were not detected on Pacemaker 2.1, so automatic HA cleanup did not run. [ QSTOR-12080 ]
* Fixed: HA pool failover failed on RHEL/AlmaLinux/Rocky 8 during SCST device removal (clustered SCSI-3 reservations are not available on el8). [ QSTOR-12182 ]
* Added: HA group creation on WD Data24 checks that IO fencing is enabled. [ QSTOR-12212 ]
* Changed: faster HA failover for pools with many volumes. [ QSTOR-12343 ]
* Fixed: HA failover could be blocked by open files, active NFS clients or a failed export, or disconnect SMB clients of other pools. [ QSTOR-12343 ]
* Fixed: in rare cases both HA nodes could import the same pool, for example after both were powered on at the same time. [ QSTOR-12343 ]
* Fixed: a volume's LUN identity could change during failover, and a suspended pool or a failed pass-through volume could hang or stop the QuantaStor service. [ QSTOR-12343 ]
* Fixed: the HA failover health check failed with a database error when an FC initiator was registered. [ QSTOR-12513 ]
* Added: an HA group option that lets a node holding fenced, suspended pools reboot itself once the peer has taken them over. [ QSTOR-12615 ]
* Changed: enabling clustered SCSI-3 reservations requires every HA group member to run 6.9.0 or newer with its installed SCST driver loaded. [ QSTOR-12630 ]
* Fixed: when several volumes are assigned to a Fibre Channel host one after another, the last one is no longer left out of the pool's ALUA device group. [ QSTOR-12697 ]
* Fixed: FC target ports are enabled only after every exported volume is mapped and its ALUA state set, so clients no longer see empty ports at startup. [ QSTOR-12724 ]
* Changed: faster, more reliable pool import and failover. [ QSTOR-12725 ]
* Fixed: a node taking over several HA pools now imports them in parallel. [ QSTOR-12728 ]
* Fixed: assigning a volume while its HA pool is exporting no longer recreates targets on the exporting node. [ QSTOR-12731 ]
* Fixed: after an HA pool owner reboots, its FC target ports return as soon as standby devices are ready instead of after 10 minutes. [ QSTOR-12765 ]
* Fixed: Windows MPIO failed to claim a QuantaStor FC node's paths after that node rebooted (event 47). [ QSTOR-12766 ]
* Fixed: an HA volume's standby device now reports the same provisioning, UNMAP and geometry as the volume. [ QSTOR-12771 ]
* Fixed: deleting an HA group left the former passive node's standby devices, ALUA groups and FC LUN maps behind. [ QSTOR-12799 ]
* Fixed: failover of an FC ALUA pool without clustered reservations now keeps clients' standby paths. [ QSTOR-12807 ]
* Fixed: an FC ALUA takeover could activate paths before the LUNs moved off the standby devices, risking lost writes. [ QSTOR-12810 ]
* Changed: the HA group dialogs have a new Advanced Settings tab, and the Tertiary System selection is removed. [ QSTOR-12861 ]
* Fixed: rebooting both HA nodes at once could panic a node during shutdown. [ QSTOR-12881 ]
* Fixed FC volume LUN numbers changing after an HA pool import and first failover, which could leave Linux initiators with standby paths to the wrong volume after the previous owner rebooted. [ QSTOR-12883 ]
* Fixed: an HA node could take up to 30 minutes to recover cluster locking, and failover during a planned node reboot could run past iSCSI/FC client timeouts. [ QSTOR-12888 ]
 
Kernel and Drivers
* Changed: Ubuntu 24.04 appliances now run kernel 6.17.0-20 with mpi3mr 8.16.1.0.0 and ixgbe 6.3.4 drivers. [ QSTOR-11647 ]
* Changed: OpenZFS updated to 2.3.9 on Ubuntu 22.04 and 24.04, fixing a write-path livelock that could block pool export. [ QSTOR-12317 ]


Licensing
Licensing
* Added logic to auto-fetch new QuantaStor License Server ca certificates during online license activation. [ QSTOR-10174 ]
* Added: license keys can be added to multiple grid systems at once. [ QSTOR-11930 ]
* Fixed: online license activation failures now report the underlying cause. [ QSTOR-12034 ]
 
Multitenancy
* Fixed: the Create Resource Group dialog closed when every subnet was already bound; the binding setting moved to the Service Configs tab. [ QSTOR-11186 ]


NAS Gateway / Cloud Containers
NAS Gateway / Cloud Containers
* Update rclone version. [ QSTOR-9512 ]
* Fixed: auto-tiering now works on the DR site after a failover. [ QSTOR-11956 ]
* Changed: removed Seagate Lyve Cloud cloud-provider support, as the service has been discontinued. [ QSTOR-12342 ]
* Fixed: added cloud-container-remove and cloud-provider-remove, cloud bucket deletion now requires --delete-cloud-bucket=true, and credential removal no longer prints the secret key. [ QSTOR-12668 ]


Network Interfaces
Network Interfaces
* Show static route parent port in the respective grid and delete dialog. [ QSTOR-10199 ]
* Fixed: a leftover cloud-init netplan file could override QuantaStor networking after qs-distupgrade. [ QSTOR-10033 ]
* Fixed an issue with onlining/restarting offline VLAN interfaces.  [ QSTOR-10180 ]
* Changed: multi-homed systems now answer ARP only on the interface that owns the address (arp_ignore=1, arp_announce=2). [ QSTOR-12324 ]
* Fixed an issue with VLAN ontop of Bonded ports on VM deployed QuantaStor instances. [ QSTOR-10180 ]
* Added: the Docker bridge and container address pools can be moved off 172.17.0.0/16 in qs_docker.conf. [ QSTOR-12325 ]
* Improve the Network bond mode descriptions, in the 'Create Bonded Port' dialog, for those that require managed switches. [ QSTOR-9842 ]
* Fixed: a static route duplicating an existing route left networking failed after reboot. [ QSTOR-12711 ]
* Added IP address place holder for DHCP interfaces that have no assigned IP address. [ QSTOR-9828 ]
* Fixed an issue with downing virtual interfaces when converting to HA virtual interface. [ QSTOR-9774 ]


Network Shares
Network Shares
* Minor Web UI fix to display share quota percent utilization. [ QSTOR-10121 ]
* Added: SFTP access to network shares. [ QSTOR-11295 ]
* Implemented a Real Time Auto Copy feature to enhance data protection for Network Shares. [ QSTOR-9929 ]
* Added: NFS and SMB client I/O statistics for network shares, per share and per client. [ QSTOR-11604 ]
* Updated Focal platform Samba to 4.18.6 backported from ubuntu upstream release. [ QSTOR-9924 ]
* Fixed: NFSv4 mounts hung after nfsdcld failed; it is now detected, alerted on and repaired automatically. [ QSTOR-11828 ]
* Show parent shares in the 'qs share-list' command results. [ QSTOR-9702 ]
* Fixed: HA pool failover did not start the NFS server on the acquiring node. [ QSTOR-11873 ]
* Fixed: network shares, share snapshots and clones could be named with leading special characters. [ QSTOR-11916 ]
* Added: the Join AD Domain dialog shows the hostname and FQDN the storage system will register in Active Directory. [ QSTOR-12105 ]
* Fixed: namespace referral lists now refresh when an interface changes or goes offline. [ QSTOR-12217 ]
* Added: a clear message when an AD user has no uidNumber/gidNumber in RFC2307 mode. [ QSTOR-12275 ]
* Changed: the namespace dialogs use a resizable share chooser that keeps selections across searches, and searches accept pool:<name>. [ QSTOR-12290 ]
* Added: share aliases and sub-shares can copy chosen settings from the parent share. [ QSTOR-12486 ]
* Added: a helper script to create and distribute the HA VIF's Kerberos identity. [ QSTOR-12574 ]
* Fixed: AD user and group searches used the previous domain settings for up to 5 minutes after a join or leave. [ QSTOR-12818 ]


Quantastor CLI
Quantastor CLI
* Update qs-iostat utility to python3. [ QSTOR-10152 ]
* Changed: qs osd-rep --disk accepts any identifier shown by qs pd-list. [ QSTOR-11920 ]
* Changed: qs osd-rep-prep finds the OSD's storage system from the OSD name, so --storage-system is optional. [ QSTOR-11921 ]
* Fixed: qs disk-list reported "disk not found" for a malformed --filter instead of a syntax error. [ QSTOR-11950 ]
* Fixed: qs --json output is now valid JSON. [ QSTOR-12003 ]
* Fixed: qs CLI commands that cannot preview now refuse --dry-run-only instead of running for real (for example disk-format, cephfs-create, site-vif-create and pool-create). [ QSTOR-12471 ]
 
REST API
* Added: an OpenAPI 3.0.3 specification for the QuantaStor REST API, for use with Swagger UI, Redoc, Postman and client generators. [ QSTOR-12675 ]


Remote Replication, Snapshots and Backup Policies
Remote Replication, Snapshots and Backup Policies
* Add the ability to force create symbolic links for backup policy outbound-move to cloud containers without a delay of 24 hours. [ QSTOR-10136 ]
* Added: new backup policies can absorb snapshots from previous backup policies. [ QSTOR-10809 ]
* Update Replication Report Summary object as failed if source system reboots right after schedule is triggered. [ QSTOR-10097 ]
* Fixed: Ceph RBD replication failed. [ QSTOR-10842 ]
* Fixed: a replication schedule could include both a parent share and its nested share, causing replication errors. [ QSTOR-11327 ]
* Added: snapshots with long-term retention now show their expiration date. [ QSTOR-11718 ]
* Fixed: some replication failures on scale-up datasets were not caught. [ QSTOR-11839 ]
* Added: a CLI command to clean up snapshots created by a replication schedule. [ QSTOR-11867 ]
* Fixed: the replication report skipped some failed replications in its total. [ QSTOR-11891 ]
* Added: snapshot schedules can create non-atomic snapshots. [ QSTOR-11913 ]
* Fixed: pool import now restores volume and share membership in snapshot and replication schedules. [ QSTOR-12115 ]
* Fixed: a scale-out replication schedule with a missing mirroring peer now fails with the reason, shown in State Detail. [ QSTOR-12693 ]
 
SCSI / NVMe Target
* Fixed: qs-util enabledualmode and disabledualmode now switch the QLogic FC driver between initiator+target and target-only mode. [ QSTOR-11791 ]
* Fixed: NVMeoF targets now always get a stable serial number derived from the volume. [ QSTOR-11851 ]
* Fixed: iSCSI ACL, host group and snapshot changes no longer reset unaffected live targets. [ QSTOR-11865 ]
* Fixed: iSCSI no longer accepts connections during boot before its targets are registered. [ QSTOR-12081 ]
* Changed: the Remove Host warning now states that existing iSCSI sessions stay connected until the initiator logs out. [ QSTOR-12518 ]
* Fixed: HPE SN1600Q and other unrecognised FC HBAs now report their vendor, and Issue FC LIP works on them. [ QSTOR-12526 ]
* Fixed: FC remote ports that left the fabric are now marked disconnected and aged out. [ QSTOR-12528 ]
* Fixed: the iSCSI service showed Running when the iSCSI target daemon was down on Ubuntu. [ QSTOR-12619 ]


SNMP
SNMP
* Regen SNMP MIB for 6.3 [ QSTOR-10228 ]
* Fixed SNMP alert traps, which had stopped being sent [ QSTOR-12890 ]
* Changed: the SNMP agent and QUANTASTOR-SYS-STATS MIB now include the object types added in 6.9; existing OIDs are unchanged. [ QSTOR-12897 ]


Security
Security
* Updated JQuery version used in the QuantaStor WebUI to 3.7.1 [ QSTOR-10064 ]
* Added: Anomaly Detection thresholds can be auto-tuned from up to 30 days of share activity. [ QSTOR-11345 ]
* Add Audit Log Viewing to the Storage System menus. [ QSTOR-10031 ]
* Added: Bitdefender integration. [ QSTOR-11768 ]
* Fixed: reserved client names such as admin are now rejected, so creating a client keyring can no longer wipe a node's Ceph credentials. [ QSTOR-11846 ]
* Fixed: CLI deletes that need multi-admin approval now report the approval status, and an approval expiration of 0 means never expire. [ QSTOR-11898 ]
* Fixed: Ceph client keyring files are now created with 0640 permissions instead of world-writable. [ QSTOR-11923 ]
* Fixed: restarting the appliance firewall no longer removes Docker rules or HA VIF block rules. [ QSTOR-11966 ]
* Fixed: a crash in the file system monitor at the end of the nightly WORM immutability scan. [ QSTOR-11979 ]
* Changed: OpenSSL updated to 3.5.7, with the FIPS 140 validated 3.1.2 module used automatically in FIPS mode. [ QSTOR-12037 ]
* Fixed: a command-injection vulnerability in the Alert Manager configuration (requires administrator access). [ QSTOR-12540 ]
* Fixed: LDAP/AD login passwords are no longer written to the audit log when the SSO authentication service is unavailable. [ QSTOR-12680 ]
* Fixed: the Ceph RGW keyring is now 0600, and open permissions left by earlier builds are corrected at startup. [ QSTOR-12702 ]
* Fixed: log bundles no longer include cephx keyrings or TLS private keys, and certificate template private keys are scrubbed. [ QSTOR-12733 ]
* Fixed: the Okta API token and other MFA secrets are no longer written to the service log. [ QSTOR-12762 ]
* Fixed: qs edr-agent-create rejects unknown module names before creating an agent. [ QSTOR-12764 ]
* Fixed: alert endpoint URLs, including ITSM API tokens, are now masked in the service and audit logs. [ QSTOR-12835 ]
 
Service Core
* Added: performance analysis tools (bpftrace, bpfcc-tools, iotop, fio, iozone, elbencho) and the qs-perftest utility are installed by default. [ QSTOR-10859 ]
* Fixed: log collection no longer prints per-task messages to the console. [ QSTOR-11035 ]
* Fixed: LUKS1-to-LUKS2 conversion of encrypted HA pools now handles HA and the peer node safely. [ QSTOR-11577 ]
* Fixed: adding or removing a site cluster node could fail intermittently. [ QSTOR-12014 ]
* Fixed: site VIF location weights were lost within about 40 seconds. [ QSTOR-12030 ]
* Fixed: qs-sendlogs failed on AlmaLinux 8 appliances. [ QSTOR-12086 ]
* Fixed: the event cache is now bounded, and clients resync when events are dropped, so deleted objects no longer linger in the web interface. [ QSTOR-12119 ]
* Fixed: the grid event cache could grow without limit on long-running grids. [ QSTOR-12120 ]
* Fixed: memory held by deleted objects is now reclaimed in the background. [ QSTOR-12122 ]
* Fixed: several quantastor.conf settings had no effect because of misspelled keys or the wrong section. [ QSTOR-12179 ]
* Fixed: renaming a volume left its NVMeoF subsystem under the old name, so later deletes failed. [ QSTOR-12210 ]
* Fixed: scrubbed log bundles now mask the user and client IP fields in qs_audit.log. [ QSTOR-12283 ]
* Fixed: two incorrect Site Cluster task and validation messages. [ QSTOR-12507 ]
* Fixed: log bundles shipped gzipped logs without scrubbing them. [ QSTOR-12634 ]


Storage Pool Management
Storage Pool Management
* Improved logic for replacing zpool UNAVAIL disks that are identified by their id instead of a device path as the disk is no longer present or visible to the OS. [ QSTOR-10129 ]
* Added: Storage System Tuning Policy management in the web interface. [ QSTOR-11952 ]
* Changed: improved SED pool cleanup in HA configurations. [ QSTOR-11963 ]
* Added: an enclosure layout view in the Storage Pool Create and Grow dialogs that previews the vdev layout by slot. [ QSTOR-11997 ]
* Added: qs pool-grow --dry-run-only reports the planned grow without changing the pool. [ QSTOR-12309 ]
* Fixed: encrypted pools could raise a false unencrypted-disk alert and go to WARNING during discovery. [ QSTOR-12592 ]
* Fixed: creating a storage pool on WD OpenFlex Data24 drives failed. [ QSTOR-12690 ]
* Fixed: Storage Pool maintenance schedules created in the Web Manager ran scrub and trim at the wrong hour, and the dialog showed the wrong Start Time. [ QSTOR-12904 ]
 
Storage Volumes
* Added: storage volumes can be resized from a new dialog. [ QSTOR-4340 ]
* Fixed: a new storage volume could show N/A utilization instead of 0% when the system also had a pass-thru volume. [ QSTOR-9076 ]
* Added: the qs-ramdisk utility creates RAM-backed SCSI disks for transport and performance testing. [ QSTOR-10222 ]
* Added: pass-through volumes can be deleted without deleting their data, and imported. [ QSTOR-11564 ]
* Changed: storage volume assign and unassign tasks now describe what changed. [ QSTOR-11737 ]
 
Universal Containers
* Added: managed Valkey, RabbitMQ and Nextcloud container services with per-service resource limits and pool-backed storage. [ QSTOR-11859 ]


Upgrade Manager and Installer
Upgrade Manager and Installer
* Automatically disable and mask uefi fwupd service on platforms that include it to prevent unplanned firmware updates. [ QSTOR-10205 ]
* Added: qs-distupgrade checks the installed QuantaStor version against the repository and notes when the installed build is newer than the repository candidate. [ QSTOR-9961 ]
* Fixed: the low boot space warning for kernel upgrades now correctly appears in the Upgrade Manager task failure message. [ QSTOR-10185 ]
* Added: qs_kernel.sh --purge-old frees space in /boot by removing kernels that are no longer needed. [ QSTOR-10176 ]
* Fixed an issue where the scst/zfs userspace packages were not being pinned on security updates. [ QSTOR-10185 ]
* Fixed: the automatic support-log upload after a failed platform upgrade could hang waiting on the terminal until it timed out. [ QSTOR-11034 ]
* Fixed an issue where qstortarget was incorrectly selected during security updates but fails due to existing pinning rules. qstortarget now correctly only gets selected for install on kernel/driver upgrades. [ QSTOR-10185 ]
* Added: the Ubuntu 24.04 (Noble) ISO is now published with its own latest-download link. [ QSTOR-11935 ]
* Added: the Ubuntu 24.04 (Noble) OVA is now published alongside the Jammy OVA. [ QSTOR-11996 ]
* Added: the qs-apt-channel utility switches QuantaStor repository channels, with validation and automatic rollback. [ QSTOR-12013 ]
* Fixed: the APT pin now covers every enabled QuantaStor update channel. [ QSTOR-12191 ]
* Fixed: apt failed on Focal appliances missing the older repository signing key. [ QSTOR-12209 ]
* Fixed: installation ISOs now include the current linux-firmware. [ QSTOR-12550 ]
* Fixed: upgrading from 6.8.x could leave a Tomcat process running and clear the firewall and Docker network rules mid-upgrade. [ QSTOR-12617 ]
* Fixed: the kernel and driver upgrade pass unloaded SCST before the reboot, leaving pools online but exporting no volumes. [ QSTOR-12636 ]
* Fixed: upgrades paused for about two minutes checking for snaps. [ QSTOR-12658 ]
* Fixed: switching repository channels could send the first upgrade pass to the Ubuntu archive; the qs-util apt channel commands now use qs-apt-channel. [ QSTOR-12809 ]
* Fixed: a multi-node kernel or driver upgrade with automatic reboots could wait indefinitely after a node rebooted if the package index could not be read. [ QSTOR-12864 ]


Web Manager
Web Manager
* Minor fix to show NA for share-alias space columns. [ QSTOR-10227 ]
* Changed: the replication and snapshot schedule dialogs use the faster share/volume chooser. [ QSTOR-10330 ]
* Login dialog shown after service restart is complete. [ QSTOR-10007 ]
* Added: online, offline, attach, detach and replace for ZFS pool devices in the web interface and CLI. [ QSTOR-11031 ]
* Fixed: System Shutdown and Restart tasks never showed Completed. [ QSTOR-11190 ]
* Changed: the Storage System "Firmware Version" field is now labeled "SCSI Target Version". [ QSTOR-11864 ]
* Added: a warning in the Storage System Shutdown and Restart dialogs when Force is selected. [ QSTOR-11888 ]
* Fixed: the SMTP password show/hide toggle in the Alert Manager email settings. [ QSTOR-11904 ]
* Fixed: target ports of an offline appliance showed Normal instead of Offline. [ QSTOR-11912 ]
* Changed: the web interface now connects through the qs-rest-soap-relay service; Tomcat is no longer required. [ QSTOR-11919 ]
* Changed: clearer storage pool type labels in the pool grid. [ QSTOR-11942 ]
* Fixed: wrong labels in the Mark Universal Storage Pool Hot-Spare dialog. [ QSTOR-11964 ]
* Changed: wider Storage System, Created and IQN columns in the storage volume sessions grid. [ QSTOR-12018 ]
* Removed: Internet Explorer 11 support from the web interface; Chrome, Firefox, Safari and current Edge are unaffected. [ QSTOR-12063 ]
* Fixed: the hostname list in Modify Ceph Object Zone Group could not scroll to the end. [ QSTOR-12101 ]
* Changed: Resource Group use cases are renamed General Purpose, Service Container Management and Object Storage Tenant, with General Purpose first. [ QSTOR-12281 ]
* Added: ZFS encryption for network shares is available again in the web interface, and encrypted shares can be replicated. [ QSTOR-12301 ]
* Dialogs that select both Storage Volumes and Network Shares now show the Network Shares tab first. [ QSTOR-12303 ]
* Navigation stack labels such as "Hosts & Portal Groups" are no longer cropped at 125% browser zoom. [ QSTOR-12313 ]
* Changed: darker dark-mode surfaces with better contrast. [ QSTOR-12314 ]
* Fixed: hard-to-read Grid dashboard tile text and toolbar selection. [ QSTOR-12326 ]
* Fixed: the enclosure diagram disappeared after zooming in the Storage Pools and Ceph OSD views. [ QSTOR-12337 ]
* Fixed: the pool Identify action opened the wrong dialog. [ QSTOR-12367 ]
* Fixed: Expand Storage Pool was offered on encrypted pools, and RAID10 offered an unsupported 4-way mirror. [ QSTOR-12368 ]
* Fixed: incorrect labels and tooltips in the Storage Pool dialogs. [ QSTOR-12369 ]
* Fixed: incorrect help text and retention tooltip in the Snapshot Schedule dialogs. [ QSTOR-12379 ]
* Fixed: the Add/Remove Pools tooltip described pool selection incorrectly. [ QSTOR-12424 ]
* Fixed: the Add/Remove Pools alert named the wrong object, and the Storage Tier utilization bar was wrong when a tier had no free space. [ QSTOR-12427 ]
* Fixed: the Grid View CPU temperature indicator colours. [ QSTOR-12452 ]
* Fixed: input validation mismatches in several dialogs. [ QSTOR-12453 ]
* Fixed: the External Dashboards toolbar did not refresh after dashboards were added or removed. [ QSTOR-12454 ]
* The dashboard time scale list now offers only the windows a view can plot, so the Trends views no longer show a range the chart is not using. [ QSTOR-12456 ]
* Fixed: the Latency chart legend showed the raw field name io_time. [ QSTOR-12457 ]
* Fixed: incorrect strings in the scale-out block dialogs. [ QSTOR-12466 ]
* The Send Support Log Files dialog is now named consistently across the toolbar, the right-click menu and the dialog title. [ QSTOR-12502 ]
* Fixed: the Getting Started dialog hid the Storage Grid Setup workflow from users without the Users & Groups tab. [ QSTOR-12511 ]
* Changed: Getting Started workflow headings are now localizable. [ QSTOR-12512 ]
* Fixed: searching for an FC WWPN in the Add Host dialog stopped all task and event updates until a page reload. [ QSTOR-12559 ]
* Fixed: a thick border around the paged Storage Volumes, Network Shares and Physical Disks grids. [ QSTOR-12650 ]
* Fixed: the Storage Pool menu and dialogs no longer offer Modify or Delete for scale-out object pool groups. [ QSTOR-12676 ]
 
=== Security Notes ===
38 security notices published since v6.8.6.015 (September 14, 2026) affect this release: 28 fixed in this release, 5 fixed by a QuantaStor security update, 5 pending a fix (12 externally-facing, 26 internally-facing). See [https://wiki.osnexus.com/security/?release=6.9.0.464 QuantaStor Security Notices for v6.9.0.464].


== v6.2.8.016 (March 5th 2024) ==
== v6.8.6.015 (September 14th 2026) ==


=== [[QuantaStor 6.2.8 Upgrade Instructions | Upgrade Instructions]] ===
=== [[QuantaStor 6.8.6 Upgrade Instructions | Upgrade Instructions]] ===
Click [[QuantaStor 6.2.8 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.2.8 Upgrade Instructions | upgrading to QuantaStor v6.2.8]].
Click [[QuantaStor 6.8.6 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.8.6 Upgrade Instructions | upgrading to QuantaStor v6.8.6]].


=== Release Notes ===
=== Release Notes ===
Alerts and Configuration Analysis
* Fixed: disk health WARNING states detected through storcli2 or SMART now raise a Disk Health alert (email/SNMP) instead of only showing in the Physical Disks view. [ QSTOR-12144 ]
Ceph Scale-out Block, Object and File
* Fixed: Scale-out Storage Cluster creation now works on routed fabrics where the selected member ports share no common subnet. [ QSTOR-12484 ]


Enclosures and Controllers
Enclosures and Controllers
* Fixed Hardware Disk to Physical disk slot correlation issues.  [ QSTOR-10135 ]
* Fixed: garbled or incomplete storcli output is now treated as unreliable instead of marking controllers and disks as missing. [ QSTOR-12118 ]
* Improved Enclosure object guid uniqueness to better support more complex/multipath setups for hardware disk to physical disk mapping. [ QSTOR-10135 ]
* Added Hardware Enclosure Serial\# property and logic to discover serial\# when available. [ QSTOR-10102 ]


Multitenancy
Network Shares
* Updated Resilio container image to 3.8.1.2250 [ QSTOR-10162 ]
* Fixed: HA VIF SMB authentication in Kerberos-only Active Directory environments. [ QSTOR-11685 ]
* Fix errors with Resilio container errors seen after upgrade. [ QSTOR-10099 ]


Remote Replication, Snapshots and Backup Policies
Remote Replication, Snapshots and Backup Policies
* Fixed an issue that could sometimes cause duplicate snapshot objects to appear for snapshot schedules with nested shares and recursive snapshot enabled. [ QSTOR-10159 ]
* Fixed the replication-schedule volume picker so it no longer hides eligible volumes whose owning node changed after an HA failover. [ QSTOR-11775 ]
* Added a check to temporarily prevent adding ZFS dataset encrypted Network Shares to remote replication schedules until the Open ZFS issue \#12014 is resolved in an upcoming ZFS release. [ QSTOR-10145 ]
* Fixed problem with backup job cancel not stopping the process. [ QSTOR-11877 ]
* Fixed an issue with using Windows Restore Previous Versions on remote Replica \_chkpnt Network Shares. [ QSTOR-10065 ]


SCSI / NVMe Target
SCSI / NVMe Target
* Updated SCST driver to 3.8.0.9038-f134517-qs01 to support latest available linux kernels for Focal platform and Focal cloud platform kernels. [ QSTOR-10151 ]
* Fixed: network ports with iSCSI/NVMe-oF disabled are never exposed as allowed iSCSI portals, including after a service restart. [ QSTOR-12330 ]
 
Storage Pool Management
* Fixed: the service no longer runs zpool status -v on the startup import path, so a pool with many data errors no longer blocks service startup. [ QSTOR-11654 ]
* Fixed: the QuantaStor service starts after multipathd so pools with many multipath devices no longer start degraded at boot. [ QSTOR-12108 ]
* Fixed: a degraded pool's missing or faulted devices now stay visible in the Storage Pool Devices view with their state instead of disappearing from the list. [ QSTOR-12109 ]


Upgrade Manager and Installer
Upgrade Manager and Installer
* Added Kernel Driver Support for QuantaStor installation on Ubuntu 20.04 images on Azure, AWS and Google Cloud. [ QSTOR-10138 ]
* Fixed: the el8 install no longer fails on the retired ipmiutil package. [ QSTOR-12257 ]
* Fixed: ISO images now carry the Ubuntu security updates for packages that reach the ISO only as dependencies. [ QSTOR-12259 ]
* Fixed: focal installs trust the current OSNEXUS repository signing key, so package upgrades from the public repository work. [ QSTOR-12274 ]
* Fixed: core-only upgrades now refresh the qstorkeyring package so repository trust updates reach every upgrade path. [ QSTOR-12280 ]
* Fixed: the install scripts trust only the current OSNEXUS repository signing keys instead of importing the published keyring wholesale. [ QSTOR-12293 ]
 
Web Manager
* Fixed: logging out of the web interface as a non-default administrator or non-admin user no longer logs a false authentication failure or raises spurious 'User Authentication Failed' alerts. [ QSTOR-11701 ]


== v6.2.7.002  (February 23rd 2024) ==
=== Security Notes ===
33 security notices published since v6.8.5.018 (August 19, 2026) affect this release: 18 fixed in this release, 12 fixed by a QuantaStor security update, 2 pending a fix, 1 pending a kernel update (18 externally-facing, 15 internally-facing). See [https://wiki.osnexus.com/security/?release=6.8.6.015 QuantaStor Security Notices for v6.8.6.015].


=== [[QuantaStor 6.2.7 Upgrade Instructions | Upgrade Instructions]] ===
== v6.8.5.018  (August 19th 2026) ==
Click [[QuantaStor 6.2.7 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.2.7 Upgrade Instructions | upgrading to QuantaStor v6.2.7]].
 
=== [[QuantaStor 6.8.5 Upgrade Instructions | Upgrade Instructions]] ===
Click [[QuantaStor 6.8.5 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.8.5 Upgrade Instructions | upgrading to QuantaStor v6.8.5]].


=== Release Notes ===
=== Release Notes ===


Licensing
Ceph Scale-out Block, Object and File
* Fixed: RGW auto-tiering rules now apply to tenant-scoped buckets; the tiering script is installed for every tenant instead of only the default one. [ QSTOR-12223 ]
* Added: auto-tiering rules can now be scoped to a tenant from the web interface and CLI - all tenants, buckets outside any tenant, or a named tenant. [ QSTOR-12225 ]
* Changed: named-tenant matching is now exact ('sc2' no longer also matches 'sc2_prod'); no rule created in the web interface could carry a named tenant before this change, so no existing configuration is affected. [ QSTOR-12225 ]
 
Enclosures and Controllers
* Fixed: Supermicro SC846 and SC847 enclosures reporting vendor SMC now match their JBOD layout profile instead of logging a layout detection failure on every discovery cycle. [ QSTOR-11947 ]
* Fixed: disks could remain displayed under a 9600-series HBA after the attached storage was removed, because a successful-but-empty discovery result was treated as a command failure and the disk list was never refreshed. [ QSTOR-12036 ]
 
Network Shares
* Fixed: Active Directory user and group lookups no longer fail intermittently where a domain controller address reverse-resolves to a name other than the controller own name. [ QSTOR-12214 ]
* Added: a domain-join option that disables reverse DNS canonicalization and sends user and group lookups to every pinned domain controller. [ QSTOR-12214 ]
* Fixed: an Active Directory user/group search now reports why it could not run, instead of always blaming a trusted-domain search filter. [ QSTOR-12216 ]
 
Service Core
* Fixed a qs_service crash under sustained DB connection pool exhaustion on large Ceph/S3 grids by serializing and capping overflow connection creation. [ QSTOR-11580 ]
* Fixed: a task that failed to resume after a service restart could block every other task from resuming. [ QSTOR-12218 ]
* Fixed: the task resume limit was not enforced for some task types, so a task that could never finish was retried on every service start. [ QSTOR-12218 ]
 
Storage Pool Management
* Added capability to simulate Corvault configurations in order to test pool configurations for redundancy. [ QSTOR-11945 ]
 
Upgrade Manager and Installer
* Fixed: a storage system upgrade task could stop resuming after the second service restart, leaving a completed upgrade reported as stuck at 30%. [ QSTOR-12177 ]
* Fixed: leftover upgrade progress files were never purged, so a finished upgrade could keep reporting stale progress. [ QSTOR-12219 ]
* Fixed: 6.8.4 ISO and OVA images were missing qstortomcat, qstorrest, qstoraws and qstorbase, which left the web interface unavailable after an offline install. The images have been rebuilt. [ QSTOR-12237 ]
* Added: an ISO build now fails if a package it declares is absent, so an incomplete image cannot ship. [ QSTOR-12237 ]


* Updated license manager certificates for online license activation. [ QSTOR-10154 ]
=== Security Notes ===
5 security notices published since v6.8.4.038 (August 13, 2026) affect this release: 5 fixed by a QuantaStor security update (3 externally-facing, 2 internally-facing). See [https://wiki.osnexus.com/security/?release=6.8.5.018 QuantaStor Security Notices for v6.8.5.018].


== v6.2.6.024 (February 8th 2024) ==
== v6.8.4.038 (August 13th 2026) ==


=== [[QuantaStor 6.2.6 Upgrade Instructions | Upgrade Instructions]] ===
=== [[QuantaStor 6.8.4 Upgrade Instructions | Upgrade Instructions]] ===
Click [[QuantaStor 6.2.6 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.2.6 Upgrade Instructions | upgrading to QuantaStor v6.2.6]].
Click [[QuantaStor 6.8.4 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.8.4 Upgrade Instructions | upgrading to QuantaStor v6.8.4]].


=== Release Notes ===
=== Release Notes ===


Enclosures and Controllers
Ceph Scale-out Block, Object and File
* Fixed an issue where the Ceph Cluster Coordinator role could become stuck on one node. The automatic fallback election methods (grid-primary, site-primary, mon-rank) were not taking effect, so when no designated coordinator was set, coordinatorship never moved (e.g. after changing the grid master). Automatic coordinator election now works correctly; explicitly designated coordinators were unaffected. [ QSTOR-12194 ]


* Added support for iofencing NVMe disks that do not support cdw11 [ QSTOR-10087 ]
Dashboards and Analytics
* Added an Alert for when a OS Software RAID1 mirror enters a degraded state. [ QSTOR-10080 ]
* Fixed the Ceph RGW dashboard on Ceph Tentacle so it correctly reports throughput/latency when transferring files to a bucket. [ QSTOR-11803 ]
* Added support LSI-3008C0 model controllers. [ QSTOR-9841 ]
* Added a check to hardware Enclosure Discovery to automatically filter out enclosures that do not have any drive slots. [ QSTOR-9716 ]


High Availability Failover
High Availability Failover
* Fixed the script to set the device reservation ability on WD Openflex Data24 [ QSTOR-12117 ]
* Fixed site cluster heartbeat rings so each ring keeps its own network address and member list. Previously the rings of a multi-ring site collapsed into a single entry with no subnet shown. [ QSTOR-12159 ]
Network Shares
* Adds an explicit domain controller to the GSSAPI ldapsearch for Active Directory user and group lookups, fixing intermittent failures on multi-controller domains. Domain joins also wait longer for a busy controller. [ QSTOR-11939 ]
* Added a cleanup-only mode to the Active Directory leave operation that reverts the local domain configuration without contacting the domain controller, for recovering from a failed or partial leave. [ QSTOR-12058 ]
* Fixed: SMB/DFS network share namespace referrals now advertise all of an appliance's reachable addresses rather than just one, so clients on any subnet can reach the share. [ QSTOR-12107 ]
* Fixed per-user share permissions so an existing entry with no recorded read/write/execute setting is treated as full access rather than no access. Modifying a share no longer rewrites ACLs when no permission actually changed. [ QSTOR-12113 ]
* Fixed Active Directory joins in SSSD-only mode against Windows Server 2025 domain controllers, which previously failed to set the machine account password. [ QSTOR-12141 ]
* Fixed placeholder values such as None, null or undefined sent by API clients being written into the Samba configuration, which stopped Samba loading the file and took SMB access down for every share on the node. [ QSTOR-12152 ]
* Fixed the Active Directory leave hanging when the PDC Emulator is offline. The leave now completes within a bounded timeout and reverts the local configuration. [ QSTOR-12161 ]
=== Security Notes ===
2 security notices published since v6.8.3.018 (August 6, 2026) affect this release: 1 fixed by a QuantaStor security update, 1 pending a fix (1 externally-facing, 1 internally-facing). See [https://wiki.osnexus.com/security/?release=6.8.4.038 QuantaStor Security Notices for v6.8.4.038].


* Fixed an issue with the I/O fencing tool where it was reporting not-supported on reservation capable nvme devices. [ QSTOR-10114 ]
== v6.8.3.018  (August 6th 2026) ==


Network Interfaces
=== [[QuantaStor 6.8.3 Upgrade Instructions | Upgrade Instructions]] ===
Click [[QuantaStor 6.8.3 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.8.3 Upgrade Instructions | upgrading to QuantaStor v6.8.3]].
 
=== Release Notes ===
 
Ceph Scale-out Block, Object and File
* Creating or removing a Ceph object storage class now commits the RGW period so it takes effect immediately. [ QSTOR-11938 ]
* A failed bucket enumeration is no longer treated as an empty list, which could silently drop bucket configuration. [ QSTOR-12130 ]
* Ceph object pool group creation now commits the RGW period, fixing persistent RGW realm/master-zone errors. [ QSTOR-12132 ]
 
Network Shares
* Fixed Active Directory join failures on domains that use older Kerberos encryption methods. [ QSTOR-11646 ]
* Changed: Active Directory joins now set 'restrict anonymous = 2' in smb.conf to block unauthenticated SID/RID enumeration. [ QSTOR-11834 ]
* Fixed: Active Directory joins no longer fail when the computer-account OU contains a space. [ QSTOR-11895 ]
* Fixed: Active Directory joins no longer disable SASL signing in the system-wide ldap.conf. [ QSTOR-11926 ]
* Added a 'Skip DNS registration' option to the Active Directory join dialog, CLI and API for sites whose DNS is managed outside AD. [ QSTOR-12083 ]
* Fixed a custom autorid Domain Range Size being discarded during an Active Directory join. [ QSTOR-12089 ]
* The Active Directory KDC field now accepts a comma-separated list of domain controllers for Kerberos failover. [ QSTOR-12091 ]
* Active Directory joins in SSSD modes no longer force user/group enumeration, which could stall startup on large domains. [ QSTOR-12092 ]
* Fixed: Active Directory joins now send the machine-account password update to the PDC Emulator, fixing 'Message stream modified' failures. [ QSTOR-12136 ]
* Fixed: leaving an Active Directory domain now updates the web UI immediately instead of still showing the old domain. [ QSTOR-12153 ]
* Changed: the Active Directory join dialog now labels its two ID-range sections distinctly so it's clear which is being edited. [ QSTOR-12154 ]
 
Service Core
* Memory optimization for grid event handling. [ QSTOR-12112 ]
 
Upgrade Manager and Installer
* The upgrade system now fetches and verifies the latest OSNEXUS package signing key before applying updates. [ QSTOR-9551 ]
* Fixed EL8 upgrades pinning a stale awscli/qstoraws package version from the RedHat repository. [ QSTOR-11991 ]
* In-place Ceph upgrades now install the shipped Ceph 20.2.2 release instead of an older pinned version. [ QSTOR-11998 ]
* Fixed a grafana-enterprise package upgrade failure that could leave dpkg in a held state during upgrades from 6.8.1. [ QSTOR-12075 ]
* Updated the Ubuntu 24.04 upgrade configuration to the shipped kernel and Ceph 20.2.2 versions. [ QSTOR-12077 ]
 
=== Security Notes ===
6 security notices published since v6.8.2.002 (July 24, 2026) affect this release: 4 fixed in this release, 1 pending a fix, 1 pending a kernel update (4 externally-facing, 2 internally-facing). See [https://wiki.osnexus.com/security/?release=6.8.3.018 QuantaStor Security Notices for v6.8.3.018].
 
== v6.8.2.002  (July 24th 2026) ==
 
=== [[QuantaStor 6.8.2 Upgrade Instructions | Upgrade Instructions]] ===
Click [[QuantaStor 6.8.2 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.8.2 Upgrade Instructions | upgrading to QuantaStor v6.8.2]].
 
=== Release Notes ===
 
Network Shares
* Fixed: AD groups could not be added to a Network Share ACL in Winbind+SSSD mode (group@DOMAIN form could not be resolved). [ QSTOR-12025 ]


* Updated static routes to use modern ip route syntax with the 'proto static' option.  [ QSTOR-10103 ]
* Updated netplan conversion to by default create static routes using ip route to match the original netplan static route rules. [ QSTOR-10103 ]


Storage Pool Management
=== Security Notes ===
17 security notices published since v6.8.1.007 (July 18, 2026) affect this release: 7 fixed in this release, 8 fixed by a QuantaStor security update, 2 pending a fix (4 externally-facing, 13 internally-facing). See [https://wiki.osnexus.com/security/?release=6.8.2.002 QuantaStor Security Notices for v6.8.2.002].


* Fixed: Activate HA Failover Group now works for Storage Pools in a degraded state missing a few devices. [ QSTOR-10100 ]
== v6.8.1.007  (July 18th 2026) ==
* Fixed: Create HA Failover Group now allows a force option if there are a small number of drives missing on the secondary node. Note that the operation will still be blocked if most/all of the drives are missing from the secondary node. [ QSTOR-10100 ]
* Added support for Zero Touch Maintenance Automatic Global Hot spare marking of Hardware RAID0 Passthrough units. [ QSTOR-9865 ]


Storage Volumes
=== [[QuantaStor 6.8.1 Upgrade Instructions | Upgrade Instructions]] ===
Click [[QuantaStor 6.8.1 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.8.1 Upgrade Instructions | upgrading to QuantaStor v6.8.1]].


* Fixed: NVMe Target Storage volumes now persist their serial numbers correctly on reboot [ QSTOR-9992 ]
=== Release Notes ===
* Fixed: the NVMe Target now correctly re-exports the allowed portals on system boot. [ QSTOR-9992 ]


Upgrade Manager and Installer
Upgrade Manager and Installer
* Fixed: platform upgrades remove the obsolete linux-tools-common package that could trigger security alerts. [ QSTOR-10314 ]
* Changed: qs-distupgrade output is timestamped and includes a pre-upgrade configuration report. [ QSTOR-11632 ]
* Added: platform distribution upgrade (qs-distupgrade) from Ubuntu 22.04 to 24.04. [ QSTOR-11651 ]
* Added: non-interactive flags for qs-distupgrade enabling fully unattended platform upgrades with automatic reboot. Use the qs-distupgrade --help to review the available usage options. [ QSTOR-11918 ]
* Fixed: platform distribution upgrades keep Ceph pinned at its current version. [ QSTOR-11955 ]
* Added: qs-distupgrade requires all Ceph nodes on the latest offered Ceph version (--skip-check-ceph-latest to override). [ QSTOR-11958 ]
* Fixed: coordinated Ceph upgrades restart all RGW daemons on hosts running multiple RGW services. [ QSTOR-11983 ]
* Fixed: a 6.8.0-only issue where service-driven package upgrades reported success without installing anything. [ QSTOR-11992 ]


* Updated mellanox-ofed-install.sh script to support the latest Focal platform 5.15.0-91-generic kernel. [ QSTOR-10117 ]
* Updated the GPG repo keys for QuantaStor RedHat / and Almalinux platform installs [ QSTOR-10107 ]
* Fixed a incorrect error log when upgrading the Kernel and Drivers on the Ubuntu Focal platform. The upgrade would succeed when the error was seen, but it was unclear to the user without full log analysis. [ QSTOR-10106 ]
* Updated upgrade manager logic to ensure recommended dependency packages are always selected for install during core service and security upgrades. [ QSTOR-10101 ]
* Fixed an issue with the qs-distupgrade script where some QuantaStor core package changes were not being applied after a platform dist-upgrade from bionic to focal. [ QSTOR-10132 ]
* Added better logging of systemctl service enablement and restart during qstorservice package installs. [ QSTOR-10134 ]


== v6.2.5.032  (January 18th 2024) ==


=== [[QuantaStor 6.2.5 Upgrade Instructions | Upgrade Instructions]] ===
 
Click [[QuantaStor 6.2.5 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.2.5 Upgrade Instructions | upgrading to QuantaStor v6.2.5]].
=== Security Notes ===
11 security notices published since v6.8.0.205 (July 8, 2026) affect this release: 6 fixed in this release, 5 fixed by a QuantaStor security update (3 externally-facing, 8 internally-facing). See [https://wiki.osnexus.com/security/?release=6.8.1.007 QuantaStor Security Notices for v6.8.1.007].
 
== v6.8.0.205  (July 8th 2026) ==
 
=== [[QuantaStor 6.8.0 Upgrade Instructions | Upgrade Instructions]] ===
Click [[QuantaStor 6.8.0 Upgrade Instructions | here]] for instructions on [[QuantaStor 6.8.0 Upgrade Instructions | upgrading to QuantaStor v6.8.0]].


=== Release Notes ===
=== Release Notes ===


Alerts and Configuration Analysis
Alerts and Configuration Analysis
* Added ssh port 22 to network connectivity checker tests. [ QSTOR-10092 ]
* Added alert webhook support for Google Chat. [ QSTOR-9793 ]
* Allow users to have a subset of alert subscriptions. [ QSTOR-11683 ]


Ceph Scale-out Block, Object and File
Ceph Scale-out Block, Object and File
* Fix bug where ceph monitor daemon is disabled. [ QSTOR-10095 ]
* Added an 'Enable Versioning' checkbox to the Create Bucket dialog (automatically enabled when object locking or governance/compliance is selected); versioning can also be enabled on existing buckets. [ QSTOR-10937 ]
* Added script to setup Ceph multi-site replication. [ QSTOR-10075 ]
* Added Ceph Tentacle 20.2.2 support to the Jammy QuantaStor ISO and package mirror. [ QSTOR-11441 ]
* QuantaStor 6.8 ships Ceph 20.2.2 (Tentacle), which includes the upstream MDS fix (tracker #76031) for the SIGABRT/crash on client eviction caused by a null MDRequestRef under sustained metadata contention. [ QSTOR-11504 ]
* The Ceph Tentacle 20.2.2 release resolves frequent radosgw (RGW) crashes related to Kafka bucket notifications. [ QSTOR-11687 ]
* Fix ceph bucket delete with large number of objects timing out. [ QSTOR-11728 ]
* Removed: the deprecated S3 bucket orphan-object cleanup feature; any existing scheduled task is disabled automatically on upgrade. [ QSTOR-11729 ]
* Fixed a regression where deleting multiple Ceph (RBD) volumes at once on a multi-node cluster would fail after the first volume. [ QSTOR-11756 ]
* Added support for major-version Ceph upgrades (through Tentacle 20.2.x) in the upgrade script. [ QSTOR-11762 ]
* Increase the timeout for "Add Node" to a ceph cluster [ QSTOR-11906 ]
 
Dashboards and Analytics
* Added: a 'qs-statsdb resetpassword' command to rotate the InfluxDB stats password; telegraf is updated in lockstep so dashboards keep working. [ QSTOR-11170 ]
* Rebuilt bundled telegraf with its vendored rclone library updated to 1.74.3, clearing CVE-2026-41176 and CVE-2026-41179. [ QSTOR-11771 ]


Enclosures and Controllers
Enclosures and Controllers
* Updated storcli version to 007.2705.0000.0000 [ QSTOR-10072 ]
* Fixed issue with "disappearing" enclosures [ QSTOR-11676 ]
* Added feature for turning off hwDisk identification for all disks, including all empty slots. [ QSTOR-10036 ]
* Added support for the Supermicro SSG-222B-NE3X24R all-NVMe server, including hardware enclosure auto-detection, enclosure layout and diagram, and NVMe drive slot mapping in the QuantaStor web management interface. [ QSTOR-11712 ]
* Added feature for turning off hwDisk identification for all disks, including all empty slots. [ QSTOR-9825 ]
 
High Availability Failover
* Add logging for manually run iofence commands [ QSTOR-11751 ]
* Fixed HA pool failover persisting a stale size=0 / isMounted=false when the pool's datasets were still mounting during fast-import; the last-known-good size and mount state are now preserved (and the corrected record wins via a bumped modified timestamp). [ QSTOR-11818 ]
* Fixed: Storage Pool HA io-fencing and failover now work over NVMe-TCP; target namespaces advertise persistent reservations. [ QSTOR-11850 ]


Kernel and Drivers
Kernel and Drivers
* Updated Linux Kernel to 5.15.0-91-generic for Focal platforms  [ QSTOR-10025 ]
* Updated Jammy platform Kernel to 6.8.0-90-generic [ QSTOR-10903 ]
* Linux Kernel 5.15.0-91-generic includes fixes for NFS v3 lockd that would cause file handles to be left open for NFSv3 clients when attempting to stop the nfs server before ZFS pool export during HA failover. [ QSTOR-10025 ]
* Updated Jammy platform ZFS driver to 2.3.5 [ QSTOR-10903 ]
* Updated ZFS to 2.1.14.qs25 for Focal platforms [ QSTOR-10025 ]
* Updated Jammy platform SCST driver to 3.10 [ QSTOR-10903 ]
* Updated Jammy platform mpi3mr driver to 8.15.1.0.0 [ QSTOR-10903 ]
* Updated the Mellanox OFED driver installer for current Ubuntu 22.04 (Jammy) kernels, fixing driver installation failures on recent kernels. [ QSTOR-11709 ]


Multitenancy
Multitenancy
* Fix an issue with missing Service Configs. [ QSTOR-10079 ]
* Added RBD namespace support, allowing multiple tenants to share a single Ceph block pool with isolated volumes. [ QSTOR-11650 ]
* Add the ability to sync nested shares when parent share is selected in Resilio. Optionally skip changing the share permissions when user is specified in the config file. [ QSTOR-10070 ]
 
NAS Gateway / Cloud Containers
* Add a fix for Samba access to Cloud Container shares. [ QSTOR-11763 ]
* Fixed: cloud containers are now accessible over SMB, not just NFS. [ QSTOR-11824 ]
* Updated the bundled AWS CLI to version 2.35.6. [ QSTOR-11827 ]
 
Network Interfaces
* Added DOCA OFED install script for jammy and upcoming noble platforms as an optional install for latest hardware. [ QSTOR-11069 ]
* Fixed the Network Interface Rename feature on systems upgraded from older releases where the 'ip' command resides in a different path. [ QSTOR-11614 ]
* Fixed: converting a local VIF to an HA VIF now preserves the original local interface if the conversion fails, so client connectivity isn't lost. [ QSTOR-11706 ]


Network Shares
Network Shares
* Block rename of nested shares. [ QSTOR-10091 ]
* Add the ability to specify per user or group permissions on Network Shares during create or modify [ QSTOR-10662 ]
* Add the ability to enable and clone a GMT snapshot [ QSTOR-4145 ]
* Fixed: intermittent 'NFS server not responding' stalls under load by raising the nfsd listen backlog from 64 to 4096. [ QSTOR-10872 ]
* Fixed: NFS client entries can now be added or removed on a share while a long-running replication or snapshot task is using it, instead of queuing. [ QSTOR-11571 ]
* Consolidated the Network Share create/modify dialog's Permission Settings into the File Masks tab — now a single 'File Masks & Permission Settings' tab containing Owner Permission Settings plus the Directory and File Create Permissions Masks. [ QSTOR-11625 ]
* Reduced excessive event logging (event spam) for network shares and related grid objects. [ QSTOR-11695 ]
* Fixed handling of CIFS share options so empty/null option values are no longer written to the Samba configuration. [ QSTOR-11731 ]
* Fixed an issue where an Active Directory group added to a share's permissions was not preserved, and excluded built-in/default AD groups from the group picker. [ QSTOR-11767 ]


Remote Replication, Snapshots and Backup Policies
Remote Replication, Snapshots and Backup Policies
* Fixed an issue where Remote Replication Schedules can fail if 2 source snapshots have the same created timestamp. [ QSTOR-10059 ]
* Copy source Network Share settings to the destination share after replication. [ QSTOR-5738 ]
* Fix a minor issue with File Age settings in the Backup Schedule Modify dialog. [ QSTOR-10056 ]
* Add support for rollback of Storage Volumes and Network Shares for a replication schedule [ QSTOR-6227 ]
* Properly transition a Replication task and Replication Report to the 'failed' status if the source system loses power during the Replication. [ QSTOR-10041 ]
* Minor updates to the auto snapshot cleanup tasks for replication schedules. [ QSTOR-11622 ]
* Add a warning if checkpoint is selected in create snapshot schedule. [ QSTOR-10038 ]
* Fix auto create of replication association after failover [ QSTOR-11700 ]
* Snapshot schedule now shows the state detail when schedule is disabled by user. [ QSTOR-11717 ]
* The Snapshot Schedule properties page now displays Retention Tags. [ QSTOR-11719 ]
* Changed: Auto-tier backup policies now display as 'Auto-Tier >> Outbound' instead of 'Move >> Outbound'. [ QSTOR-11764 ]
* Fixed the CLI replication-schedule rollback so volume and share names are accepted (not just IDs), and rolling back only volumes (or only shares) no longer also rolls back the other type. [ QSTOR-11814 ]
 
SCSI / NVMe Target
* Added: a 'Select Initiator...' picker in the Add Host dialog to choose a discovered FC remote initiator WWPN instead of typing it. [ QSTOR-10741 ]
* Renamed the sys-list 'Firmware Version' column to 'SCSI Target Version' and corrected the value to show the actual SCST target-driver version (e.g. scst-3.10.0.qs01-...) instead of the literal 'scst-SCST'. [ QSTOR-11854 ]
 
Security
* Added support for Okta multi-factor authentication. [ QSTOR-9399 ]
* Added Multi-Admin Approval support, allowing deletion of pools, buckets, shares, and volumes to require approval from a configurable number of administrators. [ QSTOR-11344 ]
* Fixed certificate update failures when using a certificate template, and ensured dependent services (such as HAProxy and RGW) are restarted after a certificate replacement. [ QSTOR-11678 ]
* Fixed an issue in the Security Manager dialog where it was showing the "password policy change" warning confirmation when no fields were changed. [ QSTOR-11708 ]
* InfluxDB (port 8086) now requires authentication and binds to localhost by default; remote access is opt-in via the tf_influxdb_remote_access.enable touchfile (authentication still enforced). [ QSTOR-11759 ]
* Added the Ubuntu security apt repositories to the default package sources and removed unused bundled components flagged by scanning (Chronograf, the LXD/core20/snapd snaps, the docker-scan plugin), clearing CVE-2024-41110, CVE-2026-34177/-34178/-34179, CVE-2020-14343, CVE-2025-68121, and CVE-2024-45337. [ QSTOR-11760 ]
* Rebuilt the bundled QuantaStor Go tools (qs_iofence, qs_pool_export_helper, qs_diskinfo) and s5cmd with a current Go toolchain, clearing CVE-2025-68121, CVE-2024-45337, CVE-2024-24790, CVE-2023-24540, and CVE-2023-24538. [ QSTOR-11761 ]
* Hardened the internal QuantaStor script service (scriptsvc): removed a command-injection vector in request handling and now enforce the service's authorization check on every call. Addresses ZDI-CAN-30036 [ QSTOR-11799 ]
* Made a number of usability and reliability improvements to the multi-admin delete-approval workflow. [ QSTOR-11804 ]
 
Service Core
* Fix tag-set where custom key/value tags were not appearing for User objects. [ QSTOR-11528 ]
* Improved Ceph OSD daemon restart handling during upgrades and reduced database load on large-scale clusters. [ QSTOR-11675 ]
* Fixed an issue with QuantaStor management service memory usage when a grid node was offline. [ QSTOR-11693 ]
* Improved crash handling and diagnostic tracing for coordinated tasks, during operations such as adding a Ceph RADOS Gateway across multiple nodes. [ QSTOR-11702 ]
* Improved logging and error checking for QuantaStor configuration database backups, including capturing error output when a backup fails. [ QSTOR-11742 ]
* Expanded diagnostic log collection to capture additional network routing details (ip route output and static-route configuration). [ QSTOR-11748 ]
* Fixed the Sessions view showing stale SMB/NFS client sessions that lingered after a reboot; the share session monitor now forces a full re-sync on its first poll after startup so sessions left over from before the reboot are cleared. [ QSTOR-11757 ]
* Fixed a self-sustaining grid event flood triggered when creating CephFS/CIFS network shares at scale: nodes no longer re-cache remote-origin (sync-applied) events, and CIFS options are diff-merged to secondaries, eliminating the event resync cascade. [ QSTOR-11785 ]
* qs-sendlogs now writes a qs_logs.runtime.info file recording the options and variables used during log collection, and reliably scrubs sensitive data from the collected database. [ QSTOR-11897 ]


Storage Volumes
Storage Pool Management
* Fixed an issue with zfs storage volumes being correctly marked as missing if the underlying zfs zvol or snapshot is removed or unavailable. [ QSTOR-10086 ]
* Added automated ordered-disk-list pool creation for Seagate Corvault and WD Data24 systems, simplifying creation of fully redundant storage pools. [ QSTOR-11538 ]
* Fixed software-encrypted pool creation so a disk block-size mismatch is detected before disks are encrypted, avoiding leftover encryption mappings. [ QSTOR-11599 ]
* ZFS deduplication now uses the faster fast_dedup mode by default, improving sustained deduplication performance. [ QSTOR-11714 ]
* When a LUKS-encrypted disk member of a pool momentarily goes offline and the hot-spare swap-in completes, if the "bad" disk in question has in the meantime come back online, its iofence status, encryption status and its related properties in its respective disk object are cleared up, making the disk available for use again. [ QSTOR-11738 ]
* Resolved a boot-time race where the OS-level zfs-import-cache.service could import ZFS pools before qs_service started. QuantaStor now imports pools with cachefile=none, masks the zfs-import-cache and zfs-import-scan services on install and upgrade, and clears the stale global ZFS cache file at boot, ensuring QuantaStor remains the sole coordinator of pool imports. [ QSTOR-11752 ]
* Add the ability to create a Storage Pool checkpoint for scale-up pools [ QSTOR-11784 ]
* Added a support-only, touchfile-gated pool-export triage call-out (qs_pool_export_triage.sh) that gathers diagnostics when a share or volume is still open during a storage pool export. Off by default (enabled via the tf_pool_export_triage.enable touchfile). [ QSTOR-11807 ]
* Resolved issue preventing HA failover functionality using SED drives. [ QSTOR-11832 ]


Upgrade Manager and Installer
Upgrade Manager and Installer
* Added a --password option to the qs\_install.sh script to allow users to provide the QuantaStor admin user password as well as the same password for the default qadmin shell account. [ QSTOR-10085 ]
* Added retry logic with a timeout to the Ceph upgrade script's daemon restarts so transient restart failures no longer abort the upgrade. [ QSTOR-10598 ]
* Fixed focal-to-jammy distribution upgrades leaving the HA cluster service (pcsd/pacemaker) inoperative due to a missing ruby-webrick dependency on Ruby 3.0+. The upgrade now installs ruby-webrick so HA failover keeps working. [ QSTOR-11841 ]
* Fixed focal-to-jammy distribution upgrades that could abort and roll back when Ubuntu's unattended-upgrades/apt-daily held the apt lock or needrestart blocked on a prompt. QuantaStor now disables automatic apt upgrades and runs the distribution upgrade non-interactively. [ QSTOR-11874 ]
* Fixed a Noble (24.04) upgrade failure where qs-upgrade aborted because the dpkg package version string could not be parsed. QuantaStor now compares Debian package versions using dpkg's native version ordering. [ QSTOR-11881 ]


Web Manager
Web Manager
* Clear the replica summary graph when there are no report entries. [ QSTOR-10057 ]
* Minor fix to the "Copy to Clipboard" for task descriptions [ QSTOR-9574 ]
* Fix enclosure graph rendering issue where a selected system with multiple valid controllers was not being render because of one underlying MD BOOT Controller. [ QSTOR-10055 ]
* Added a method field and a sortable timestamp column to the web interface Audit Log Analyzer. [ QSTOR-11191 ]
* Updated web interface translations across the major supported languages. [ QSTOR-11585 ]
* Added a 'Create Share' option to the share template menu to quickly create a share applying all of a template's settings. [ QSTOR-11595 ]
* Fixed the web interface to display all configured Ceph MDS daemons instead of limiting the list to five. [ QSTOR-11615 ]
* Add auto select convenience buttons to the Storage Volume multi-delete dialog. [ QSTOR-11707 ]
 
 
=== Security Notes ===
12 security notices published since v6.7.6.001 (June 25, 2026) affect this release: 3 fixed in this release, 9 fixed by a QuantaStor security update (3 externally-facing, 9 internally-facing). See [https://wiki.osnexus.com/security/?release=6.8.0.205 QuantaStor Security Notices for v6.8.0.205].


= Archived Change Logs =
= Archived Change Logs =
== [[Change Log Archive 2026]] ==
== [[Change Log Archive 2025]] ==
== [[Change Log Archive 2024]] ==


== [[Change Log Archive 2023]] ==
== [[Change Log Archive 2023]] ==

Latest revision as of 16:31, 17 November 2026

Change Log Overview

The change log contains a detailed summary of the changes made for each new release of QuantaStor. Each change made to QuantaStor is tracked with an engineering ticket number in the form [QSTOR-NNNN]. This information is provided so that if you're working with OSNexus support you can track the availability of a given product feature add, improvement, or maintenance fix to a specific release. For information on how to upgrade your storage system please see the Upgrade Guide.

Versioning System

QuantaStor version numbers consist of four (4) parts. A major number (M), minor number (N), maintenance update number (U) and build number (B) of the form M.N.U.B such as 6.0.2.006.

QuantaStor upgrades go directly from whatever version you are running to the latest version, with no interim steps so for example if you are at v6.0.1 you will upgrade directly to the latest version which may be v6.1.2

Product Update Types

Major Updates (X.y.z)

If the first digit in the version number has changed this indicates a major product release such as QuantaStor v5 to QuantaStor v6. Major releases generally come with a new underlying Linux platform as the basis. For example, the standard ISO install for QuantaStor v5 is based on Ubuntu 18.04 whereas QuantaStor v6 is based on Ubuntu 20.04. Major product releases generally include new features and maintenance updates and are supported on multiple underlying Linux platforms. For example, QuantaStor v6 is available as an upgrade on Ubuntu 18.04 and Ubuntu 20.04 based systems as well as RHEL and RHEL derivative platforms. Platform upgrades to move systems forward (eg 18.04 to 20.04) are made generally available in the first feature update release, such as QuantaStor 6.1.

Feature Updates (x.Y.z)

If the second digit in the version number has changed this indicates a feature release such as 6.1 going to 6.2. Feature releases add new product features as well as a maintenance updates. No additional licensing is required to access any new features that added but note that QuantaStor licenses do not include licenses for 3rd party products and services that may integrate with QuantaStor.

Maintenance Updates (x.y.Z)

If the third digit of the version number has changed this such as 6.0.2 or 6.0.3 it represents a maintenance update to address one or more support tickets. These updates generally do not contain new features, only fixes to address specific issues. Maintenance releases generally ship every 4 to 8 weeks. All maintenance "aka hot-fix" releases are provided against the current stable release so upgrading to the latest QuantaStor version is generally required to apply them.

Build Number (x.y.z.BBB)

The build number and corresponding build reference hash can be largely ignored. The build number simply increments with each commit that is made to the source tree. The hash is provided to make it easier for OSNEXUS support and engineering to quickly confirm difference between your current running release and any more modern release.

Boot Install Image and other downloads

ISO/DVD/USB Boot Install Image

NOTE: Archived older versions of QuantaStor ISO install media and CLI tools can be found here.

Recent Change Logs

New Security Notices

No security notices published since the latest release, v7.0.1.147 (November 17, 2026), affect QuantaStor. Last checked November 17, 2026.

v7.0.1.147 (November 17th 2026)

Upgrade Instructions

Click here for instructions on upgrading to QuantaStor v7.0.1.

Release Notes

Alerts and Configuration Analysis

  • Fixed: Pacemaker node-down and node-up alerts not reaching QuantaStor on Ubuntu 24.04. [ QSTOR-12546 ]
  • Fixed share quota Warning and Urgent alerts not being raised again when another share entered a free-space band that had already alerted. [ QSTOR-12827 ]
  • Fixed: alert-config-set --pause-alerts accepting invalid pause periods and unknown alert types. [ QSTOR-12838 ]

Ceph Scale-out Block, Object and File

  • Fixed S3 gateways that stopped answering requests while their service was still running not being detected or alerted. [ QSTOR-10646 ]
  • Fixed: a failed Ceph cluster create no longer leaves leftovers that block every later create on those nodes. [ QSTOR-12224 ]
  • Fixed: Object Pool Group creation failing when the Ceph cluster already had the default RGW zone group. [ QSTOR-12271 ]
  • Fixed: S3 requests by hostname failing with NoSuchBucket on nodes with no DNS search domain. [ QSTOR-12749 ]
  • Fixed: cloud storage classes created in the web interface now write tiered objects to the destination bucket you choose. [ QSTOR-12857 ]
  • Added: a Ceph-over-ZFS Optimized tuning profile. [ QSTOR-12869 ]
  • Fixed: encrypted Ceph OSDs failing to start after a reboot that followed an OSD encryption key replace. [ QSTOR-12914 ]
  • Fixed: the Ubuntu 22.04 to 24.04 platform upgrade on Ceph cluster nodes marking OSDs out while the node was upgraded. [ QSTOR-12917 ]
  • Fixed a new Ceph RBD volume sometimes being mapped twice on the node that created it, which made deleting the volume fail with 'image has watchers'. [ QSTOR-12990 ]
  • Fixed a force-deleted Ceph pool leaving its RBD volumes mapped on every node. [ QSTOR-12991 ]
  • Fixed: the HAProxy load balancer not working on Ceph cluster members without a local RADOS gateway. [ QSTOR-13066 ]
  • Fixed: encrypted Ceph OSDs failing authentication after a restart following a cephx key conversion. [ QSTOR-13126 ]
  • Fixed: S3 auto-tiering for tenant buckets stopping after a coordinated Ceph upgrade from Reef to 20.2.x. [ QSTOR-13141 ]
  • New Ceph clusters no longer enable discard on every OSD, and existing clusters have it turned off without restarting OSDs. [ QSTOR-13172 ]
  • Delete Storage Class now opens on the right-clicked storage class, and STANDARD no longer offers Delete. [ QSTOR-13221 ]

Dashboards and Analytics

  • Fixed: performance dashboards showing no data or wrong values for NFS and SMB shares and Ceph pools, OSDs, MDS and RGW. [ QSTOR-13128 ]
  • Fixed: install-ceph-prometheus now runs on QuantaStor 7.0 (Ubuntu 24.04) and installs current Prometheus releases. [ QSTOR-13315 ]

Enclosures and Controllers

  • Fixed: a long-running disk identify blocking the request to turn the light off, and ignoring cancel. [ QSTOR-12785 ]
  • Fixed: disk identify failing to blink drives on some enclosures. [ QSTOR-12844 ]
  • Fixed: a failed disk identify blink now reports an error instead of Completed. [ QSTOR-12845 ]
  • Added: an option to create Corvault disk groups and pools interleaved. [ QSTOR-12955 ]
  • Fixed: enclosure auto-detection never selecting the Supermicro SC946 60-bay JBOD layouts. [ QSTOR-12966 ]
  • Fixed: External System Profile Apply ignoring the Assign to Hosts setting. [ QSTOR-13068 ]
  • Fixed: NVMe drives in the Supermicro ASG-2115S-NE332R getting no drive slot, so Identify could not light their bays. [ QSTOR-13103 ]

High Availability Failover

  • Fixed a node placed in manual standby returning to Active when another cluster node's name starts with the same name. [ QSTOR-11917 ]
  • Creating a site virtual interface right after creating its site cluster now succeeds, and a failed create no longer leaves the address active. [ QSTOR-12473 ]
  • Unmanaged virtual interface messages and refused manual failovers now say when the site cluster is in maintenance mode. [ QSTOR-12475 ]
  • Fixed a storage pool that had stopped writing to disk being reported healthy with no alert. [ QSTOR-12667 ]
  • Fixed: when a pool export overran the HA group's Export Timeout, failover imported the pool without first fencing the exporting node. [ QSTOR-12671 ]
  • Fixed: a storage pool import reporting Completed when the pool could not be imported. [ QSTOR-12674 ]
  • Fixed: a service restart or one HA pool's failover moving another HA pool off its healthy node. [ QSTOR-12763 ]
  • Fixed: stopping HA cluster services on a node giving up after 15 seconds and leaving the services in an inconsistent state. [ QSTOR-12834 ]
  • Fixed package updates on Ubuntu 24.04 restarting HA cluster services, which could cause an unplanned HA pool failover. [ QSTOR-12913 ]
  • Fixed site-vif-delete --convert-to-vif briefly dropping the IP address and raising HA group maintenance mode alerts. [ QSTOR-12931 ]
  • HA nodes no longer report a failed lock manager service at every boot. [ QSTOR-13021 ]
  • Fixed: adding a second cluster ring to an HA pair with clustered persistent reservations no longer hangs the peer node. [ QSTOR-13115 ]
  • Fixed restarting Pacemaker on an HA node with clustered SCSI3-PR reservations taking minutes and leaving the cluster lock service failed. [ QSTOR-13175 ]
  • Fixed: corosync and pacemaker restarting within a minute of an administrator's 'pcs cluster stop'. [ QSTOR-13227 ]

NAS Gateway / Cloud Containers

  • Fixed: importing a cloud container onto an already-mounted export path leaving it without an NFS export or SMB share. [ QSTOR-12919 ]

Network Interfaces

  • Fixed: network ports no longer get renamed (eno1 to eno1np0) by kernel and distribution upgrades, which dropped IP, bond and HA VIF settings. [ QSTOR-11072 ]
  • Creating a bonded network interface no longer leaves a second, conflicting definition for the same bond in the network configuration. [ QSTOR-12097 ]
  • Fixed: an empty DNS search suffix breaking the host name and AD join, which now checks DNS identity and domain controller ports. [ QSTOR-13112 ]

Network Shares

  • Kerberos NFS (sec=krb5) users on AD-joined systems now keep their AD identity instead of nobody. [ QSTOR-8198 ]
  • Network share search is now case-insensitive, also matches share descriptions, and runs on Enter or as you type. [ QSTOR-10789 ]
  • Added: AD join can use a pre-created computer account. [ QSTOR-11334 ]
  • Fixed a share whose dataset mount hangs being retried every hour, leaking stuck processes and blocking other share operations. [ QSTOR-12664 ]
  • Fixed a storage system removed from a network share namespace continuing to advertise the namespace over SMB and NFS until another namespace change. [ QSTOR-12921 ]
  • Fixed: Changing a network share from the CLI no longer resets its owner group unless --share-owner-group is given. [ QSTOR-12945 ]
  • Fixed share rename leaving a share disabled when an NFS client had a file open, and accepting conflicting names. [ QSTOR-12993 ]
  • Fixed: AD join with a wrong administrator password now says the password is wrong. [ QSTOR-12999 ]
  • Fixed copying share settings removing SMB user access from the source share, which could leave it open to all users. [ QSTOR-13176 ]
  • Fixed: Network share owners and groups can again be set to a numeric UID or GID that has no matching user or group on the system. [ QSTOR-13320 ]

Quantastor CLI

  • Fixed sys-modify silently ignoring --domain-suffix, and DNS server changes resetting a customized host FQDN in /etc/hosts. [ QSTOR-12116 ]
  • Fixed: CLI options given as --option=false acting as if they were set to true. [ QSTOR-12782 ]
  • Fixed: qs rg-create failing when --scope-type is left at its default. [ QSTOR-12916 ]
  • Fixed: qs rg-res-add rejecting Object Tenants. [ QSTOR-12918 ]

Remote Replication, Snapshots and Backup Policies

  • Replication schedules that reference a removed storage system link now switch to the current link instead of failing every run. [ QSTOR-11735 ]
  • Fixed: a replication schedule rollback issued right after a replication run reporting Completed without restoring anything. [ QSTOR-12880 ]
  • Fixed backup policies losing their destination sub-folder on modify and jobs not reporting failures or cancels correctly. [ QSTOR-12886 ]

SCSI / NVMe Target

  • Added: Fibre Channel ports now report their negotiated and supported speeds. [ QSTOR-12855 ]
  • Fixed: a Portal Group with no portals exposing its iSCSI volumes on every portal. [ QSTOR-12996 ]
  • Added: an alert listing the iSCSI sessions closed when a Portal Group change removes a portal. [ QSTOR-13018 ]
  • Fixed NVMe/TCP volumes created on a running pool not being offered to hosts until the pool restarted or failed over. [ QSTOR-13153 ]

SNMP

  • Fixed: the SNMP agent not reporting new objects in a table that was empty when the agent started. [ QSTOR-12905 ]

Security

  • Assigning Okta MFA to a user with no email is refused with a clear error. [ QSTOR-12885 ]
  • Fixed: multi-admin approval requests not expiring on the node that created them. [ QSTOR-12981 ]
  • Updated the gSOAP library to 2.8.146, the Genivia critical security release. [ QSTOR-13088 ]
  • Added: an optional mitigation for CVE-2026-98432 (qs-util ipoptfilter enable). [ QSTOR-13148 ]

Server Integration

  • Added: a QuantaStor storage plugin for Proxmox VE with snapshots, clones, migration, backups and multipath on iSCSI volumes. [ QSTOR-10624 ]
  • Changed: the Proxmox VE plugin is now a native storage type installed from a .deb package and survives Proxmox upgrades. [ QSTOR-11575 ]
  • Added LXC container support to the QuantaStor Proxmox VE plugin: container root disks can be placed on QuantaStor iSCSI block volumes. [ QSTOR-11968 ]
  • Documented the guest I/O pause during a QuantaStor HA pool failover (about 30-45 seconds, no errors) and the iSCSI timeout guidance. [ QSTOR-12311 ]
  • Documented that storage entries from the Proxmox plugin pre-release stop working on 1.0.0 and later. [ QSTOR-12487 ]
  • Added API token authentication: a QuantaStor application token can be used in place of a password. [ QSTOR-12489 ]
  • Fixed: Proxmox snapshot deletes failing with err=493 and leaving the guest locked. [ QSTOR-12537 ]
  • Released Proxmox VE plugin 1.0.1. [ QSTOR-13004 ]
  • Added: iSCSI multipath support in the Proxmox VE plugin. [ QSTOR-13005 ]
  • Fixed: Proxmox container backups in snapshot mode failing on QuantaStor storage. [ QSTOR-13046 ]
  • Added: full clones of a Proxmox VM from a snapshot, and fixed disk creation failing with "already exists". [ QSTOR-13047 ]

Service Core

  • Fixed: Send Support Logs reporting Completed when a slow collection or upload never delivered the bundle. [ QSTOR-12703 ]
  • Fixed support bundles leaving out configuration changes made shortly before log collection. [ QSTOR-12705 ]
  • Added the Recovery Manager: generate, download, upload and restore storage system configuration bundles from the web UI and CLI. [ QSTOR-13010 ]
  • Fixed the NFS/SMB statistics collector logging sudo entries to the system log every 5 seconds. [ QSTOR-13011 ]

Storage Pool Management

  • A storage pool is no longer marked offline when a disk-space query fails or times out while the pool is still imported and serving data. [ QSTOR-12804 ]
  • Fixed: a serving storage pool being marked unmounted after a failed size check. [ QSTOR-13149 ]

Universal Containers

  • A universal container service whose port is already in use now fails to start with a clear error, instead of repeatedly restarting Docker. [ QSTOR-12329 ]
  • Fixed: a Service Container port conflict restarting Docker on the node. [ QSTOR-12898 ]

Upgrade Manager and Installer

  • Added: SHA-256 and SHA-512 checksum files for each ISO, OVA and OVA zip download, alongside the existing MD5. [ QSTOR-11155 ]
  • Changed: the Ubuntu 22.04 to 24.04 platform upgrade now requires QuantaStor 6.8.0 or later. [ QSTOR-12624 ]
  • The Ceph upgrade now rejects an invalid target Ceph version with a clear error instead of ignoring it. [ QSTOR-12627 ]
  • Fixed: Ceph upgrades failing or hanging on grids that reach the package repository through an HTTP/HTTPS proxy. [ QSTOR-12670 ]
  • Added: qs-upgrade --cephPackagesOnly upgrades the Ceph packages on a system that is not part of a Ceph cluster. [ QSTOR-12775 ]
  • Fixed: the coordinated Ceph upgrade on Ubuntu 24.04 restarting every Ceph daemon at once during the package stage. [ QSTOR-12851 ]
  • Fixed: the coordinated Ceph upgrade reporting Completed when the post-upgrade cleanup failed. [ QSTOR-13146 ]
  • Fixed: a coordinated Ceph upgrade refused with "Detected current upgrade in progress" after an earlier command-line upgrade. [ QSTOR-13163 ]
  • Fixed: upgrades not refreshing the sudoers rules used by the NFS/SMB statistics collector and Pacemaker node alerts. [ QSTOR-13164 ]

Web Manager

  • Long property values such as descriptions now wrap in the web UI property views instead of being cut off. [ QSTOR-9385 ]
  • Fixed: ribbon bar buttons no longer draw borders in dark mode. [ QSTOR-9562 ]
  • Fixed: adding a node to a running site cluster now completes and the node appears in the web UI. [ QSTOR-12005 ]
  • Fixed: Modify Grid Time Zone Settings preselecting Europe/Andorra on UTC grids, and Etc/UTC can now be selected. [ QSTOR-12519 ]
  • Fixed: tree searches now accept the same pool: and comma terms as the grids, and Physical Disks search finds disks. [ QSTOR-12819 ]
  • Fixed % Utilized progress bars flashing white while grids refresh in dark mode. [ QSTOR-13222 ]

Security Notes

19 security notices published since v6.9.2.001 (November 2, 2026) affect this release: 13 fixed in this release, 2 fixed by a QuantaStor security update, 3 pending a fix, 1 pending a kernel update (13 externally-facing, 6 internally-facing). See QuantaStor Security Notices for v7.0.1.147.

v6.9.2.001 (November 2nd 2026)

Upgrade Instructions

Click here for instructions on upgrading to QuantaStor v6.9.2.

Release Notes

Remote Replication, Snapshots and Backup Policies

  • Fixed: snapshot and replication schedules could stop running after upgrading to 6.9.1 on systems with many snapshots. [ QSTOR-13162 ]

Security Notes

No security notices published since v6.9.1.001 (October 28, 2026) affect this release.

v6.9.1.001 (October 28th 2026)

Upgrade Instructions

Click here for instructions on upgrading to QuantaStor v6.9.1.

Release Notes

Security

  • Fixed: QuantaStor failed to start in FIPS mode after upgrading to 6.9 when a 3DES-encrypted license manager certificate was present. [ QSTOR-13041 ]
  • Fixed: QuantaStor now distrusts the compromised eMudhra and Comodo AAA root CAs (USN-8944-1); please upgrade with security updates included (Core + Security, or qs-upgrade without --coreOnly) to install the change. [ QSTOR-13070 ]

Security Notes

10 security notices published since v6.9.0.464 (October 15, 2026) affect this release: 8 fixed in this release, 1 fixed by a QuantaStor security update, 1 pending a fix (5 externally-facing, 5 internally-facing). See QuantaStor Security Notices for v6.9.1.001.

v6.9.0.464 (October 15th 2026)

Upgrade Instructions

Click here for instructions on upgrading to QuantaStor v6.9.0.

Release Notes

Alerts and Configuration Analysis

  • Added: a Zabbix alert handler (Zabbix 7.0+) with an importable template of triggers and dashboards. [ QSTOR-10802 ]
  • Added: alerts can be paused. [ QSTOR-11665 ]
  • Added: alerts for scale-out share capacity thresholds. [ QSTOR-11911 ]
  • Fixed: report schedules did not save to a custom local save location. [ QSTOR-11937 ]
  • Fixed: a removed disk could keep raising a Disk Health Non-optimal alert with a blank device and serial. [ QSTOR-12230 ]
  • Changed: Discord is no longer listed by default in the Alert Manager ITSM integrations; it can be re-enabled in qs_alerthandlers.conf. [ QSTOR-12341 ]

Ceph Scale-out Block, Object and File

  • Added: replication of scale-out (CephFS) network shares to a second Ceph cluster using CephFS snapshot mirroring. [ QSTOR-9709 ]
  • Added: bucket object lifecycle policies for external S3 endpoints. [ QSTOR-11603 ]
  • Fixed: client keyrings for volumes on erasure-coded pools now include the data pool, so direct RBD mapping works. [ QSTOR-11610 ]
  • Fixed: RADOS gateway logs are now rotated daily and capped at 100 MB so they cannot fill the boot drive. [ QSTOR-11645 ]
  • Added: per-bucket rate limits on operations and bandwidth for object storage. [ QSTOR-11836 ]
  • Fixed: deleting a scale-out object pool from Storage Management now points to qs object-pool-group-delete instead of failing with an unclear error. [ QSTOR-11896 ]
  • Fixed: retrying a failed Ceph node add no longer stops with "key does not match". [ QSTOR-11907 ]
  • Fixed: qs cckr-assoc-get failed with a database column error. [ QSTOR-11924 ]
  • Fixed: deleting a Ceph pool or CephFS share now removes it from client keyrings that referenced it. [ QSTOR-11925 ]
  • Fixed: cloning a Ceph RBD volume or snapshot failed. [ QSTOR-11928 ]
  • Fixed: a failed Ceph RBD clone can now be retried, and force-deleting a Ceph pool removes its volume records. [ QSTOR-11929 ]
  • Fixed: qs osc-create rejected some recommended storage class names, such as EXPRESS_ONEZONE and OUTPOSTS. [ QSTOR-11943 ]
  • Fixed: load-balanced object storage now redirects ports 80 and 443 to the load balancer's ports instead of falling back to a single gateway. [ QSTOR-12073 ]
  • Fixed: qs object-user-create --tenant failed with "Unable to locate tenant". [ QSTOR-12123 ]
  • Changed: bucket and S3 user discovery now uses the RGW admin API, so large namespaces scan much faster and a failed gateway reply no longer removes buckets or users. [ QSTOR-12133 ]
  • Fixed: qs bucket-get now finds tenant-scoped buckets by bare name or tenant/bucket. [ QSTOR-12148 ]
  • Changed: Ceph updated to Tentacle 20.2.4 on Ubuntu 22.04 and 24.04. [ QSTOR-12162 ]
  • Fixed: a deleted object storage class data pool was recreated by the gateway; deleting the default STANDARD pool is now refused. [ QSTOR-12165 ]
  • Added: Ceph over ZFS - zvols in a ZFS storage pool (including dRAID) can be used as Ceph OSD devices. [ QSTOR-12166 ]
  • Added: storage classes can be deleted. [ QSTOR-12178 ]
  • Fixed: upgrading RGW multisite deployments to Ceph 20.2.4 keeps zones replicating, and secure SigV4 is restored once every zone is upgraded (CVE-2026-54330). [ QSTOR-12269 ]
  • Changed: Ceph 20.2.4 clusters keep legacy CephX keys so kernel clients keep working; qs ceph-cluster-key-convert converts daemon keys to aes256k (CVE-2025-30156). [ QSTOR-12270 ]
  • Fixed: RGW admin operations now use HTTPS when the gateway runs in HTTPS mode. [ QSTOR-12294 ]
  • Changed: Create Ceph Cluster reports progress per stage, and the Create OSDs dialog keeps selections when the search filter changes. [ QSTOR-12323 ]
  • Fixed: Ceph OSD creation now takes seconds instead of minutes on large devices. [ QSTOR-12332 ]
  • Changed: the default metadata DB size for new Ceph OSDs is sized from the selected devices, up to 3% of raw capacity. [ QSTOR-12461 ]
  • Added: cephfs-mirror and rbd-mirror are installed on Ubuntu 22.04 and 24.04 appliances. [ QSTOR-12547 ]
  • Fixed: Ceph OSDs on a loopback-alias public address showed Offline, and a missing OSD now raises an alert. [ QSTOR-12734 ]
  • Fixed: qs-s3util move-objects now rejects a storage class that does not exist (override with --skip-class-check). [ QSTOR-12755 ]
  • Fixed Replace OSD losing encryption and DB/WAL devices; the replacement disk is now cleared automatically and disks in use are refused. [ QSTOR-12894 ]
  • Fixed encrypted Ceph OSD key replace failing on grids with journal devices, CLI OSD create dropping DB/WAL devices when no size was given, and failed OSD creates leaving orphaned OSD IDs. [ QSTOR-12901 ]

Dashboards and Analytics

  • Added: External Dashboards - custom dashboards embedded in the Web Manager and organized into groups. [ QSTOR-9308 ]
  • Added: a View Report option to view pool reports in the web interface. [ QSTOR-9707 ]
  • Added: a network share usage tracker and dashboard, filterable by user or group. [ QSTOR-11720 ]
  • Added: a Replication view on the Share and Volume dashboards that charts replication transfer rate. [ QSTOR-12289 ]
  • Fixed: dashboard tile actions are now keyboard accessible and their colours meet WCAG AA contrast. [ QSTOR-12491 ]
  • Changed: Trends charts draw lines instead of filled areas, legends toggle series again, and the pool chart plots used space against capacity. [ QSTOR-12575 ]
  • Fixed: volume capacity and trend statistics were not recorded when optimized stats collection was enabled. [ QSTOR-12576 ]
  • Fixed: Trends and Usage windows of 12 hours or longer showed zero capacity. [ QSTOR-12577 ]
  • Fixed: updated statistics rollup settings were not applied to upgraded systems. [ QSTOR-12578 ]

Disk Management

  • Changed: removed unnecessary SMART query log messages on CORVAULT systems. [ QSTOR-11583 ]
  • Added: a [tag:VALUE] disk filter, so tagged disks can be selected by tag, for example in pool-create --disk-list. [ QSTOR-11887 ]
  • Added: a Mode option for the disk performance test to run disks serially, in parallel, or grouped by vdev or enclosure. [ QSTOR-11953 ]
  • Fixed: healthy SATA drives behind a SAS HBA raised false non-optimal health alerts. [ QSTOR-12082 ]
  • Fixed: drives behind an HPE Smart Array Gen10 HBA (smartpqi) were not shown on Ubuntu 24.04. [ QSTOR-12726 ]

Enclosures and Controllers

  • Fixed: a Seagate CORVAULT that was not fully populated was given the wrong enclosure layout, or none. [ QSTOR-12088 ]
  • Added support for the Dell PowerEdge R740xd 24-bay NVMe server, including its enclosure layout and drive slot mapping in Hardware Enclosures & Controllers. [ QSTOR-12944 ]

High Availability Failover

  • Validated: Windows Server 2022 Hyper-V failover clusters with Cluster Shared Volumes on QuantaStor Fibre Channel HA pools. [ QSTOR-10677 ]
  • Added: clustered SCSI-3 persistent reservations for HA storage pools, so client reservations (Hyper-V, Windows Server Failover Clustering) survive failover. [ QSTOR-10705 ]
  • Fixed: service SPNs on an HA VIF's AD computer object now keep working after failover. [ QSTOR-10884 ]
  • Fixed: moving a site cluster VIF to a node with no location constraint now fails at once with an explanation instead of after a minute of retries. [ QSTOR-11605 ]
  • Fixed: deleting an HA VIF from an active pool could fail the pool over to the standby node. [ QSTOR-11694 ]
  • Fixed: Set Grid Sync Master reverted within seconds on grids with a Grid Management VIF. [ QSTOR-11698 ]
  • Fixed: the FC link-state failover policy now applies only to pools that export volumes over FC. [ QSTOR-11855 ]
  • Fixed: an HA node lost its manual-standby state after a reboot or distribution upgrade. [ QSTOR-11957 ]
  • Fixed: deleting an HA pool or HA VIF could trigger a spurious failover on the peer. [ QSTOR-12002 ]
  • Fixed: a stale site-cluster association could produce a corosync configuration that stopped corosync from starting. [ QSTOR-12026 ]
  • Fixed: a manual-standby setting left over from an earlier configuration could force HA nodes into standby after an upgrade. [ QSTOR-12044 ]
  • Fixed: failed resource actions were not detected on Pacemaker 2.1, so automatic HA cleanup did not run. [ QSTOR-12080 ]
  • Fixed: HA pool failover failed on RHEL/AlmaLinux/Rocky 8 during SCST device removal (clustered SCSI-3 reservations are not available on el8). [ QSTOR-12182 ]
  • Added: HA group creation on WD Data24 checks that IO fencing is enabled. [ QSTOR-12212 ]
  • Changed: faster HA failover for pools with many volumes. [ QSTOR-12343 ]
  • Fixed: HA failover could be blocked by open files, active NFS clients or a failed export, or disconnect SMB clients of other pools. [ QSTOR-12343 ]
  • Fixed: in rare cases both HA nodes could import the same pool, for example after both were powered on at the same time. [ QSTOR-12343 ]
  • Fixed: a volume's LUN identity could change during failover, and a suspended pool or a failed pass-through volume could hang or stop the QuantaStor service. [ QSTOR-12343 ]
  • Fixed: the HA failover health check failed with a database error when an FC initiator was registered. [ QSTOR-12513 ]
  • Added: an HA group option that lets a node holding fenced, suspended pools reboot itself once the peer has taken them over. [ QSTOR-12615 ]
  • Changed: enabling clustered SCSI-3 reservations requires every HA group member to run 6.9.0 or newer with its installed SCST driver loaded. [ QSTOR-12630 ]
  • Fixed: when several volumes are assigned to a Fibre Channel host one after another, the last one is no longer left out of the pool's ALUA device group. [ QSTOR-12697 ]
  • Fixed: FC target ports are enabled only after every exported volume is mapped and its ALUA state set, so clients no longer see empty ports at startup. [ QSTOR-12724 ]
  • Changed: faster, more reliable pool import and failover. [ QSTOR-12725 ]
  • Fixed: a node taking over several HA pools now imports them in parallel. [ QSTOR-12728 ]
  • Fixed: assigning a volume while its HA pool is exporting no longer recreates targets on the exporting node. [ QSTOR-12731 ]
  • Fixed: after an HA pool owner reboots, its FC target ports return as soon as standby devices are ready instead of after 10 minutes. [ QSTOR-12765 ]
  • Fixed: Windows MPIO failed to claim a QuantaStor FC node's paths after that node rebooted (event 47). [ QSTOR-12766 ]
  • Fixed: an HA volume's standby device now reports the same provisioning, UNMAP and geometry as the volume. [ QSTOR-12771 ]
  • Fixed: deleting an HA group left the former passive node's standby devices, ALUA groups and FC LUN maps behind. [ QSTOR-12799 ]
  • Fixed: failover of an FC ALUA pool without clustered reservations now keeps clients' standby paths. [ QSTOR-12807 ]
  • Fixed: an FC ALUA takeover could activate paths before the LUNs moved off the standby devices, risking lost writes. [ QSTOR-12810 ]
  • Changed: the HA group dialogs have a new Advanced Settings tab, and the Tertiary System selection is removed. [ QSTOR-12861 ]
  • Fixed: rebooting both HA nodes at once could panic a node during shutdown. [ QSTOR-12881 ]
  • Fixed FC volume LUN numbers changing after an HA pool import and first failover, which could leave Linux initiators with standby paths to the wrong volume after the previous owner rebooted. [ QSTOR-12883 ]
  • Fixed: an HA node could take up to 30 minutes to recover cluster locking, and failover during a planned node reboot could run past iSCSI/FC client timeouts. [ QSTOR-12888 ]

Kernel and Drivers

  • Changed: Ubuntu 24.04 appliances now run kernel 6.17.0-20 with mpi3mr 8.16.1.0.0 and ixgbe 6.3.4 drivers. [ QSTOR-11647 ]
  • Changed: OpenZFS updated to 2.3.9 on Ubuntu 22.04 and 24.04, fixing a write-path livelock that could block pool export. [ QSTOR-12317 ]

Licensing

  • Added: license keys can be added to multiple grid systems at once. [ QSTOR-11930 ]
  • Fixed: online license activation failures now report the underlying cause. [ QSTOR-12034 ]

Multitenancy

  • Fixed: the Create Resource Group dialog closed when every subnet was already bound; the binding setting moved to the Service Configs tab. [ QSTOR-11186 ]

NAS Gateway / Cloud Containers

  • Fixed: auto-tiering now works on the DR site after a failover. [ QSTOR-11956 ]
  • Changed: removed Seagate Lyve Cloud cloud-provider support, as the service has been discontinued. [ QSTOR-12342 ]
  • Fixed: added cloud-container-remove and cloud-provider-remove, cloud bucket deletion now requires --delete-cloud-bucket=true, and credential removal no longer prints the secret key. [ QSTOR-12668 ]

Network Interfaces

  • Fixed: a leftover cloud-init netplan file could override QuantaStor networking after qs-distupgrade. [ QSTOR-10033 ]
  • Changed: multi-homed systems now answer ARP only on the interface that owns the address (arp_ignore=1, arp_announce=2). [ QSTOR-12324 ]
  • Added: the Docker bridge and container address pools can be moved off 172.17.0.0/16 in qs_docker.conf. [ QSTOR-12325 ]
  • Fixed: a static route duplicating an existing route left networking failed after reboot. [ QSTOR-12711 ]

Network Shares

  • Added: SFTP access to network shares. [ QSTOR-11295 ]
  • Added: NFS and SMB client I/O statistics for network shares, per share and per client. [ QSTOR-11604 ]
  • Fixed: NFSv4 mounts hung after nfsdcld failed; it is now detected, alerted on and repaired automatically. [ QSTOR-11828 ]
  • Fixed: HA pool failover did not start the NFS server on the acquiring node. [ QSTOR-11873 ]
  • Fixed: network shares, share snapshots and clones could be named with leading special characters. [ QSTOR-11916 ]
  • Added: the Join AD Domain dialog shows the hostname and FQDN the storage system will register in Active Directory. [ QSTOR-12105 ]
  • Fixed: namespace referral lists now refresh when an interface changes or goes offline. [ QSTOR-12217 ]
  • Added: a clear message when an AD user has no uidNumber/gidNumber in RFC2307 mode. [ QSTOR-12275 ]
  • Changed: the namespace dialogs use a resizable share chooser that keeps selections across searches, and searches accept pool:<name>. [ QSTOR-12290 ]
  • Added: share aliases and sub-shares can copy chosen settings from the parent share. [ QSTOR-12486 ]
  • Added: a helper script to create and distribute the HA VIF's Kerberos identity. [ QSTOR-12574 ]
  • Fixed: AD user and group searches used the previous domain settings for up to 5 minutes after a join or leave. [ QSTOR-12818 ]

Quantastor CLI

  • Changed: qs osd-rep --disk accepts any identifier shown by qs pd-list. [ QSTOR-11920 ]
  • Changed: qs osd-rep-prep finds the OSD's storage system from the OSD name, so --storage-system is optional. [ QSTOR-11921 ]
  • Fixed: qs disk-list reported "disk not found" for a malformed --filter instead of a syntax error. [ QSTOR-11950 ]
  • Fixed: qs --json output is now valid JSON. [ QSTOR-12003 ]
  • Fixed: qs CLI commands that cannot preview now refuse --dry-run-only instead of running for real (for example disk-format, cephfs-create, site-vif-create and pool-create). [ QSTOR-12471 ]

REST API

  • Added: an OpenAPI 3.0.3 specification for the QuantaStor REST API, for use with Swagger UI, Redoc, Postman and client generators. [ QSTOR-12675 ]

Remote Replication, Snapshots and Backup Policies

  • Added: new backup policies can absorb snapshots from previous backup policies. [ QSTOR-10809 ]
  • Fixed: Ceph RBD replication failed. [ QSTOR-10842 ]
  • Fixed: a replication schedule could include both a parent share and its nested share, causing replication errors. [ QSTOR-11327 ]
  • Added: snapshots with long-term retention now show their expiration date. [ QSTOR-11718 ]
  • Fixed: some replication failures on scale-up datasets were not caught. [ QSTOR-11839 ]
  • Added: a CLI command to clean up snapshots created by a replication schedule. [ QSTOR-11867 ]
  • Fixed: the replication report skipped some failed replications in its total. [ QSTOR-11891 ]
  • Added: snapshot schedules can create non-atomic snapshots. [ QSTOR-11913 ]
  • Fixed: pool import now restores volume and share membership in snapshot and replication schedules. [ QSTOR-12115 ]
  • Fixed: a scale-out replication schedule with a missing mirroring peer now fails with the reason, shown in State Detail. [ QSTOR-12693 ]

SCSI / NVMe Target

  • Fixed: qs-util enabledualmode and disabledualmode now switch the QLogic FC driver between initiator+target and target-only mode. [ QSTOR-11791 ]
  • Fixed: NVMeoF targets now always get a stable serial number derived from the volume. [ QSTOR-11851 ]
  • Fixed: iSCSI ACL, host group and snapshot changes no longer reset unaffected live targets. [ QSTOR-11865 ]
  • Fixed: iSCSI no longer accepts connections during boot before its targets are registered. [ QSTOR-12081 ]
  • Changed: the Remove Host warning now states that existing iSCSI sessions stay connected until the initiator logs out. [ QSTOR-12518 ]
  • Fixed: HPE SN1600Q and other unrecognised FC HBAs now report their vendor, and Issue FC LIP works on them. [ QSTOR-12526 ]
  • Fixed: FC remote ports that left the fabric are now marked disconnected and aged out. [ QSTOR-12528 ]
  • Fixed: the iSCSI service showed Running when the iSCSI target daemon was down on Ubuntu. [ QSTOR-12619 ]

SNMP

  • Fixed SNMP alert traps, which had stopped being sent [ QSTOR-12890 ]
  • Changed: the SNMP agent and QUANTASTOR-SYS-STATS MIB now include the object types added in 6.9; existing OIDs are unchanged. [ QSTOR-12897 ]

Security

  • Added: Anomaly Detection thresholds can be auto-tuned from up to 30 days of share activity. [ QSTOR-11345 ]
  • Added: Bitdefender integration. [ QSTOR-11768 ]
  • Fixed: reserved client names such as admin are now rejected, so creating a client keyring can no longer wipe a node's Ceph credentials. [ QSTOR-11846 ]
  • Fixed: CLI deletes that need multi-admin approval now report the approval status, and an approval expiration of 0 means never expire. [ QSTOR-11898 ]
  • Fixed: Ceph client keyring files are now created with 0640 permissions instead of world-writable. [ QSTOR-11923 ]
  • Fixed: restarting the appliance firewall no longer removes Docker rules or HA VIF block rules. [ QSTOR-11966 ]
  • Fixed: a crash in the file system monitor at the end of the nightly WORM immutability scan. [ QSTOR-11979 ]
  • Changed: OpenSSL updated to 3.5.7, with the FIPS 140 validated 3.1.2 module used automatically in FIPS mode. [ QSTOR-12037 ]
  • Fixed: a command-injection vulnerability in the Alert Manager configuration (requires administrator access). [ QSTOR-12540 ]
  • Fixed: LDAP/AD login passwords are no longer written to the audit log when the SSO authentication service is unavailable. [ QSTOR-12680 ]
  • Fixed: the Ceph RGW keyring is now 0600, and open permissions left by earlier builds are corrected at startup. [ QSTOR-12702 ]
  • Fixed: log bundles no longer include cephx keyrings or TLS private keys, and certificate template private keys are scrubbed. [ QSTOR-12733 ]
  • Fixed: the Okta API token and other MFA secrets are no longer written to the service log. [ QSTOR-12762 ]
  • Fixed: qs edr-agent-create rejects unknown module names before creating an agent. [ QSTOR-12764 ]
  • Fixed: alert endpoint URLs, including ITSM API tokens, are now masked in the service and audit logs. [ QSTOR-12835 ]

Service Core

  • Added: performance analysis tools (bpftrace, bpfcc-tools, iotop, fio, iozone, elbencho) and the qs-perftest utility are installed by default. [ QSTOR-10859 ]
  • Fixed: log collection no longer prints per-task messages to the console. [ QSTOR-11035 ]
  • Fixed: LUKS1-to-LUKS2 conversion of encrypted HA pools now handles HA and the peer node safely. [ QSTOR-11577 ]
  • Fixed: adding or removing a site cluster node could fail intermittently. [ QSTOR-12014 ]
  • Fixed: site VIF location weights were lost within about 40 seconds. [ QSTOR-12030 ]
  • Fixed: qs-sendlogs failed on AlmaLinux 8 appliances. [ QSTOR-12086 ]
  • Fixed: the event cache is now bounded, and clients resync when events are dropped, so deleted objects no longer linger in the web interface. [ QSTOR-12119 ]
  • Fixed: the grid event cache could grow without limit on long-running grids. [ QSTOR-12120 ]
  • Fixed: memory held by deleted objects is now reclaimed in the background. [ QSTOR-12122 ]
  • Fixed: several quantastor.conf settings had no effect because of misspelled keys or the wrong section. [ QSTOR-12179 ]
  • Fixed: renaming a volume left its NVMeoF subsystem under the old name, so later deletes failed. [ QSTOR-12210 ]
  • Fixed: scrubbed log bundles now mask the user and client IP fields in qs_audit.log. [ QSTOR-12283 ]
  • Fixed: two incorrect Site Cluster task and validation messages. [ QSTOR-12507 ]
  • Fixed: log bundles shipped gzipped logs without scrubbing them. [ QSTOR-12634 ]

Storage Pool Management

  • Added: Storage System Tuning Policy management in the web interface. [ QSTOR-11952 ]
  • Changed: improved SED pool cleanup in HA configurations. [ QSTOR-11963 ]
  • Added: an enclosure layout view in the Storage Pool Create and Grow dialogs that previews the vdev layout by slot. [ QSTOR-11997 ]
  • Added: qs pool-grow --dry-run-only reports the planned grow without changing the pool. [ QSTOR-12309 ]
  • Fixed: encrypted pools could raise a false unencrypted-disk alert and go to WARNING during discovery. [ QSTOR-12592 ]
  • Fixed: creating a storage pool on WD OpenFlex Data24 drives failed. [ QSTOR-12690 ]
  • Fixed: Storage Pool maintenance schedules created in the Web Manager ran scrub and trim at the wrong hour, and the dialog showed the wrong Start Time. [ QSTOR-12904 ]

Storage Volumes

  • Added: storage volumes can be resized from a new dialog. [ QSTOR-4340 ]
  • Fixed: a new storage volume could show N/A utilization instead of 0% when the system also had a pass-thru volume. [ QSTOR-9076 ]
  • Added: the qs-ramdisk utility creates RAM-backed SCSI disks for transport and performance testing. [ QSTOR-10222 ]
  • Added: pass-through volumes can be deleted without deleting their data, and imported. [ QSTOR-11564 ]
  • Changed: storage volume assign and unassign tasks now describe what changed. [ QSTOR-11737 ]

Universal Containers

  • Added: managed Valkey, RabbitMQ and Nextcloud container services with per-service resource limits and pool-backed storage. [ QSTOR-11859 ]

Upgrade Manager and Installer

  • Added: qs-distupgrade checks the installed QuantaStor version against the repository and notes when the installed build is newer than the repository candidate. [ QSTOR-9961 ]
  • Added: qs_kernel.sh --purge-old frees space in /boot by removing kernels that are no longer needed. [ QSTOR-10176 ]
  • Fixed: the automatic support-log upload after a failed platform upgrade could hang waiting on the terminal until it timed out. [ QSTOR-11034 ]
  • Added: the Ubuntu 24.04 (Noble) ISO is now published with its own latest-download link. [ QSTOR-11935 ]
  • Added: the Ubuntu 24.04 (Noble) OVA is now published alongside the Jammy OVA. [ QSTOR-11996 ]
  • Added: the qs-apt-channel utility switches QuantaStor repository channels, with validation and automatic rollback. [ QSTOR-12013 ]
  • Fixed: the APT pin now covers every enabled QuantaStor update channel. [ QSTOR-12191 ]
  • Fixed: apt failed on Focal appliances missing the older repository signing key. [ QSTOR-12209 ]
  • Fixed: installation ISOs now include the current linux-firmware. [ QSTOR-12550 ]
  • Fixed: upgrading from 6.8.x could leave a Tomcat process running and clear the firewall and Docker network rules mid-upgrade. [ QSTOR-12617 ]
  • Fixed: the kernel and driver upgrade pass unloaded SCST before the reboot, leaving pools online but exporting no volumes. [ QSTOR-12636 ]
  • Fixed: upgrades paused for about two minutes checking for snaps. [ QSTOR-12658 ]
  • Fixed: switching repository channels could send the first upgrade pass to the Ubuntu archive; the qs-util apt channel commands now use qs-apt-channel. [ QSTOR-12809 ]
  • Fixed: a multi-node kernel or driver upgrade with automatic reboots could wait indefinitely after a node rebooted if the package index could not be read. [ QSTOR-12864 ]

Web Manager

  • Changed: the replication and snapshot schedule dialogs use the faster share/volume chooser. [ QSTOR-10330 ]
  • Added: online, offline, attach, detach and replace for ZFS pool devices in the web interface and CLI. [ QSTOR-11031 ]
  • Fixed: System Shutdown and Restart tasks never showed Completed. [ QSTOR-11190 ]
  • Changed: the Storage System "Firmware Version" field is now labeled "SCSI Target Version". [ QSTOR-11864 ]
  • Added: a warning in the Storage System Shutdown and Restart dialogs when Force is selected. [ QSTOR-11888 ]
  • Fixed: the SMTP password show/hide toggle in the Alert Manager email settings. [ QSTOR-11904 ]
  • Fixed: target ports of an offline appliance showed Normal instead of Offline. [ QSTOR-11912 ]
  • Changed: the web interface now connects through the qs-rest-soap-relay service; Tomcat is no longer required. [ QSTOR-11919 ]
  • Changed: clearer storage pool type labels in the pool grid. [ QSTOR-11942 ]
  • Fixed: wrong labels in the Mark Universal Storage Pool Hot-Spare dialog. [ QSTOR-11964 ]
  • Changed: wider Storage System, Created and IQN columns in the storage volume sessions grid. [ QSTOR-12018 ]
  • Removed: Internet Explorer 11 support from the web interface; Chrome, Firefox, Safari and current Edge are unaffected. [ QSTOR-12063 ]
  • Fixed: the hostname list in Modify Ceph Object Zone Group could not scroll to the end. [ QSTOR-12101 ]
  • Changed: Resource Group use cases are renamed General Purpose, Service Container Management and Object Storage Tenant, with General Purpose first. [ QSTOR-12281 ]
  • Added: ZFS encryption for network shares is available again in the web interface, and encrypted shares can be replicated. [ QSTOR-12301 ]
  • Dialogs that select both Storage Volumes and Network Shares now show the Network Shares tab first. [ QSTOR-12303 ]
  • Navigation stack labels such as "Hosts & Portal Groups" are no longer cropped at 125% browser zoom. [ QSTOR-12313 ]
  • Changed: darker dark-mode surfaces with better contrast. [ QSTOR-12314 ]
  • Fixed: hard-to-read Grid dashboard tile text and toolbar selection. [ QSTOR-12326 ]
  • Fixed: the enclosure diagram disappeared after zooming in the Storage Pools and Ceph OSD views. [ QSTOR-12337 ]
  • Fixed: the pool Identify action opened the wrong dialog. [ QSTOR-12367 ]
  • Fixed: Expand Storage Pool was offered on encrypted pools, and RAID10 offered an unsupported 4-way mirror. [ QSTOR-12368 ]
  • Fixed: incorrect labels and tooltips in the Storage Pool dialogs. [ QSTOR-12369 ]
  • Fixed: incorrect help text and retention tooltip in the Snapshot Schedule dialogs. [ QSTOR-12379 ]
  • Fixed: the Add/Remove Pools tooltip described pool selection incorrectly. [ QSTOR-12424 ]
  • Fixed: the Add/Remove Pools alert named the wrong object, and the Storage Tier utilization bar was wrong when a tier had no free space. [ QSTOR-12427 ]
  • Fixed: the Grid View CPU temperature indicator colours. [ QSTOR-12452 ]
  • Fixed: input validation mismatches in several dialogs. [ QSTOR-12453 ]
  • Fixed: the External Dashboards toolbar did not refresh after dashboards were added or removed. [ QSTOR-12454 ]
  • The dashboard time scale list now offers only the windows a view can plot, so the Trends views no longer show a range the chart is not using. [ QSTOR-12456 ]
  • Fixed: the Latency chart legend showed the raw field name io_time. [ QSTOR-12457 ]
  • Fixed: incorrect strings in the scale-out block dialogs. [ QSTOR-12466 ]
  • The Send Support Log Files dialog is now named consistently across the toolbar, the right-click menu and the dialog title. [ QSTOR-12502 ]
  • Fixed: the Getting Started dialog hid the Storage Grid Setup workflow from users without the Users & Groups tab. [ QSTOR-12511 ]
  • Changed: Getting Started workflow headings are now localizable. [ QSTOR-12512 ]
  • Fixed: searching for an FC WWPN in the Add Host dialog stopped all task and event updates until a page reload. [ QSTOR-12559 ]
  • Fixed: a thick border around the paged Storage Volumes, Network Shares and Physical Disks grids. [ QSTOR-12650 ]
  • Fixed: the Storage Pool menu and dialogs no longer offer Modify or Delete for scale-out object pool groups. [ QSTOR-12676 ]

Security Notes

38 security notices published since v6.8.6.015 (September 14, 2026) affect this release: 28 fixed in this release, 5 fixed by a QuantaStor security update, 5 pending a fix (12 externally-facing, 26 internally-facing). See QuantaStor Security Notices for v6.9.0.464.

v6.8.6.015 (September 14th 2026)

Upgrade Instructions

Click here for instructions on upgrading to QuantaStor v6.8.6.

Release Notes

Alerts and Configuration Analysis

  • Fixed: disk health WARNING states detected through storcli2 or SMART now raise a Disk Health alert (email/SNMP) instead of only showing in the Physical Disks view. [ QSTOR-12144 ]

Ceph Scale-out Block, Object and File

  • Fixed: Scale-out Storage Cluster creation now works on routed fabrics where the selected member ports share no common subnet. [ QSTOR-12484 ]

Enclosures and Controllers

  • Fixed: garbled or incomplete storcli output is now treated as unreliable instead of marking controllers and disks as missing. [ QSTOR-12118 ]

Network Shares

  • Fixed: HA VIF SMB authentication in Kerberos-only Active Directory environments. [ QSTOR-11685 ]

Remote Replication, Snapshots and Backup Policies

  • Fixed the replication-schedule volume picker so it no longer hides eligible volumes whose owning node changed after an HA failover. [ QSTOR-11775 ]
  • Fixed problem with backup job cancel not stopping the process. [ QSTOR-11877 ]

SCSI / NVMe Target

  • Fixed: network ports with iSCSI/NVMe-oF disabled are never exposed as allowed iSCSI portals, including after a service restart. [ QSTOR-12330 ]

Storage Pool Management

  • Fixed: the service no longer runs zpool status -v on the startup import path, so a pool with many data errors no longer blocks service startup. [ QSTOR-11654 ]
  • Fixed: the QuantaStor service starts after multipathd so pools with many multipath devices no longer start degraded at boot. [ QSTOR-12108 ]
  • Fixed: a degraded pool's missing or faulted devices now stay visible in the Storage Pool Devices view with their state instead of disappearing from the list. [ QSTOR-12109 ]

Upgrade Manager and Installer

  • Fixed: the el8 install no longer fails on the retired ipmiutil package. [ QSTOR-12257 ]
  • Fixed: ISO images now carry the Ubuntu security updates for packages that reach the ISO only as dependencies. [ QSTOR-12259 ]
  • Fixed: focal installs trust the current OSNEXUS repository signing key, so package upgrades from the public repository work. [ QSTOR-12274 ]
  • Fixed: core-only upgrades now refresh the qstorkeyring package so repository trust updates reach every upgrade path. [ QSTOR-12280 ]
  • Fixed: the install scripts trust only the current OSNEXUS repository signing keys instead of importing the published keyring wholesale. [ QSTOR-12293 ]

Web Manager

  • Fixed: logging out of the web interface as a non-default administrator or non-admin user no longer logs a false authentication failure or raises spurious 'User Authentication Failed' alerts. [ QSTOR-11701 ]

Security Notes

33 security notices published since v6.8.5.018 (August 19, 2026) affect this release: 18 fixed in this release, 12 fixed by a QuantaStor security update, 2 pending a fix, 1 pending a kernel update (18 externally-facing, 15 internally-facing). See QuantaStor Security Notices for v6.8.6.015.

v6.8.5.018 (August 19th 2026)

Upgrade Instructions

Click here for instructions on upgrading to QuantaStor v6.8.5.

Release Notes

Ceph Scale-out Block, Object and File

  • Fixed: RGW auto-tiering rules now apply to tenant-scoped buckets; the tiering script is installed for every tenant instead of only the default one. [ QSTOR-12223 ]
  • Added: auto-tiering rules can now be scoped to a tenant from the web interface and CLI - all tenants, buckets outside any tenant, or a named tenant. [ QSTOR-12225 ]
  • Changed: named-tenant matching is now exact ('sc2' no longer also matches 'sc2_prod'); no rule created in the web interface could carry a named tenant before this change, so no existing configuration is affected. [ QSTOR-12225 ]

Enclosures and Controllers

  • Fixed: Supermicro SC846 and SC847 enclosures reporting vendor SMC now match their JBOD layout profile instead of logging a layout detection failure on every discovery cycle. [ QSTOR-11947 ]
  • Fixed: disks could remain displayed under a 9600-series HBA after the attached storage was removed, because a successful-but-empty discovery result was treated as a command failure and the disk list was never refreshed. [ QSTOR-12036 ]

Network Shares

  • Fixed: Active Directory user and group lookups no longer fail intermittently where a domain controller address reverse-resolves to a name other than the controller own name. [ QSTOR-12214 ]
  • Added: a domain-join option that disables reverse DNS canonicalization and sends user and group lookups to every pinned domain controller. [ QSTOR-12214 ]
  • Fixed: an Active Directory user/group search now reports why it could not run, instead of always blaming a trusted-domain search filter. [ QSTOR-12216 ]

Service Core

  • Fixed a qs_service crash under sustained DB connection pool exhaustion on large Ceph/S3 grids by serializing and capping overflow connection creation. [ QSTOR-11580 ]
  • Fixed: a task that failed to resume after a service restart could block every other task from resuming. [ QSTOR-12218 ]
  • Fixed: the task resume limit was not enforced for some task types, so a task that could never finish was retried on every service start. [ QSTOR-12218 ]

Storage Pool Management

  • Added capability to simulate Corvault configurations in order to test pool configurations for redundancy. [ QSTOR-11945 ]

Upgrade Manager and Installer

  • Fixed: a storage system upgrade task could stop resuming after the second service restart, leaving a completed upgrade reported as stuck at 30%. [ QSTOR-12177 ]
  • Fixed: leftover upgrade progress files were never purged, so a finished upgrade could keep reporting stale progress. [ QSTOR-12219 ]
  • Fixed: 6.8.4 ISO and OVA images were missing qstortomcat, qstorrest, qstoraws and qstorbase, which left the web interface unavailable after an offline install. The images have been rebuilt. [ QSTOR-12237 ]
  • Added: an ISO build now fails if a package it declares is absent, so an incomplete image cannot ship. [ QSTOR-12237 ]

Security Notes

5 security notices published since v6.8.4.038 (August 13, 2026) affect this release: 5 fixed by a QuantaStor security update (3 externally-facing, 2 internally-facing). See QuantaStor Security Notices for v6.8.5.018.

v6.8.4.038 (August 13th 2026)

Upgrade Instructions

Click here for instructions on upgrading to QuantaStor v6.8.4.

Release Notes

Ceph Scale-out Block, Object and File

  • Fixed an issue where the Ceph Cluster Coordinator role could become stuck on one node. The automatic fallback election methods (grid-primary, site-primary, mon-rank) were not taking effect, so when no designated coordinator was set, coordinatorship never moved (e.g. after changing the grid master). Automatic coordinator election now works correctly; explicitly designated coordinators were unaffected. [ QSTOR-12194 ]

Dashboards and Analytics

  • Fixed the Ceph RGW dashboard on Ceph Tentacle so it correctly reports throughput/latency when transferring files to a bucket. [ QSTOR-11803 ]

High Availability Failover

  • Fixed the script to set the device reservation ability on WD Openflex Data24 [ QSTOR-12117 ]
  • Fixed site cluster heartbeat rings so each ring keeps its own network address and member list. Previously the rings of a multi-ring site collapsed into a single entry with no subnet shown. [ QSTOR-12159 ]

Network Shares

  • Adds an explicit domain controller to the GSSAPI ldapsearch for Active Directory user and group lookups, fixing intermittent failures on multi-controller domains. Domain joins also wait longer for a busy controller. [ QSTOR-11939 ]
  • Added a cleanup-only mode to the Active Directory leave operation that reverts the local domain configuration without contacting the domain controller, for recovering from a failed or partial leave. [ QSTOR-12058 ]
  • Fixed: SMB/DFS network share namespace referrals now advertise all of an appliance's reachable addresses rather than just one, so clients on any subnet can reach the share. [ QSTOR-12107 ]
  • Fixed per-user share permissions so an existing entry with no recorded read/write/execute setting is treated as full access rather than no access. Modifying a share no longer rewrites ACLs when no permission actually changed. [ QSTOR-12113 ]
  • Fixed Active Directory joins in SSSD-only mode against Windows Server 2025 domain controllers, which previously failed to set the machine account password. [ QSTOR-12141 ]
  • Fixed placeholder values such as None, null or undefined sent by API clients being written into the Samba configuration, which stopped Samba loading the file and took SMB access down for every share on the node. [ QSTOR-12152 ]
  • Fixed the Active Directory leave hanging when the PDC Emulator is offline. The leave now completes within a bounded timeout and reverts the local configuration. [ QSTOR-12161 ]

Security Notes

2 security notices published since v6.8.3.018 (August 6, 2026) affect this release: 1 fixed by a QuantaStor security update, 1 pending a fix (1 externally-facing, 1 internally-facing). See QuantaStor Security Notices for v6.8.4.038.

v6.8.3.018 (August 6th 2026)

Upgrade Instructions

Click here for instructions on upgrading to QuantaStor v6.8.3.

Release Notes

Ceph Scale-out Block, Object and File

  • Creating or removing a Ceph object storage class now commits the RGW period so it takes effect immediately. [ QSTOR-11938 ]
  • A failed bucket enumeration is no longer treated as an empty list, which could silently drop bucket configuration. [ QSTOR-12130 ]
  • Ceph object pool group creation now commits the RGW period, fixing persistent RGW realm/master-zone errors. [ QSTOR-12132 ]

Network Shares

  • Fixed Active Directory join failures on domains that use older Kerberos encryption methods. [ QSTOR-11646 ]
  • Changed: Active Directory joins now set 'restrict anonymous = 2' in smb.conf to block unauthenticated SID/RID enumeration. [ QSTOR-11834 ]
  • Fixed: Active Directory joins no longer fail when the computer-account OU contains a space. [ QSTOR-11895 ]
  • Fixed: Active Directory joins no longer disable SASL signing in the system-wide ldap.conf. [ QSTOR-11926 ]
  • Added a 'Skip DNS registration' option to the Active Directory join dialog, CLI and API for sites whose DNS is managed outside AD. [ QSTOR-12083 ]
  • Fixed a custom autorid Domain Range Size being discarded during an Active Directory join. [ QSTOR-12089 ]
  • The Active Directory KDC field now accepts a comma-separated list of domain controllers for Kerberos failover. [ QSTOR-12091 ]
  • Active Directory joins in SSSD modes no longer force user/group enumeration, which could stall startup on large domains. [ QSTOR-12092 ]
  • Fixed: Active Directory joins now send the machine-account password update to the PDC Emulator, fixing 'Message stream modified' failures. [ QSTOR-12136 ]
  • Fixed: leaving an Active Directory domain now updates the web UI immediately instead of still showing the old domain. [ QSTOR-12153 ]
  • Changed: the Active Directory join dialog now labels its two ID-range sections distinctly so it's clear which is being edited. [ QSTOR-12154 ]

Service Core

  • Memory optimization for grid event handling. [ QSTOR-12112 ]

Upgrade Manager and Installer

  • The upgrade system now fetches and verifies the latest OSNEXUS package signing key before applying updates. [ QSTOR-9551 ]
  • Fixed EL8 upgrades pinning a stale awscli/qstoraws package version from the RedHat repository. [ QSTOR-11991 ]
  • In-place Ceph upgrades now install the shipped Ceph 20.2.2 release instead of an older pinned version. [ QSTOR-11998 ]
  • Fixed a grafana-enterprise package upgrade failure that could leave dpkg in a held state during upgrades from 6.8.1. [ QSTOR-12075 ]
  • Updated the Ubuntu 24.04 upgrade configuration to the shipped kernel and Ceph 20.2.2 versions. [ QSTOR-12077 ]

Security Notes

6 security notices published since v6.8.2.002 (July 24, 2026) affect this release: 4 fixed in this release, 1 pending a fix, 1 pending a kernel update (4 externally-facing, 2 internally-facing). See QuantaStor Security Notices for v6.8.3.018.

v6.8.2.002 (July 24th 2026)

Upgrade Instructions

Click here for instructions on upgrading to QuantaStor v6.8.2.

Release Notes

Network Shares

  • Fixed: AD groups could not be added to a Network Share ACL in Winbind+SSSD mode (group@DOMAIN form could not be resolved). [ QSTOR-12025 ]


Security Notes

17 security notices published since v6.8.1.007 (July 18, 2026) affect this release: 7 fixed in this release, 8 fixed by a QuantaStor security update, 2 pending a fix (4 externally-facing, 13 internally-facing). See QuantaStor Security Notices for v6.8.2.002.

v6.8.1.007 (July 18th 2026)

Upgrade Instructions

Click here for instructions on upgrading to QuantaStor v6.8.1.

Release Notes

Upgrade Manager and Installer

  • Fixed: platform upgrades remove the obsolete linux-tools-common package that could trigger security alerts. [ QSTOR-10314 ]
  • Changed: qs-distupgrade output is timestamped and includes a pre-upgrade configuration report. [ QSTOR-11632 ]
  • Added: platform distribution upgrade (qs-distupgrade) from Ubuntu 22.04 to 24.04. [ QSTOR-11651 ]
  • Added: non-interactive flags for qs-distupgrade enabling fully unattended platform upgrades with automatic reboot. Use the qs-distupgrade --help to review the available usage options. [ QSTOR-11918 ]
  • Fixed: platform distribution upgrades keep Ceph pinned at its current version. [ QSTOR-11955 ]
  • Added: qs-distupgrade requires all Ceph nodes on the latest offered Ceph version (--skip-check-ceph-latest to override). [ QSTOR-11958 ]
  • Fixed: coordinated Ceph upgrades restart all RGW daemons on hosts running multiple RGW services. [ QSTOR-11983 ]
  • Fixed: a 6.8.0-only issue where service-driven package upgrades reported success without installing anything. [ QSTOR-11992 ]




Security Notes

11 security notices published since v6.8.0.205 (July 8, 2026) affect this release: 6 fixed in this release, 5 fixed by a QuantaStor security update (3 externally-facing, 8 internally-facing). See QuantaStor Security Notices for v6.8.1.007.

v6.8.0.205 (July 8th 2026)

Upgrade Instructions

Click here for instructions on upgrading to QuantaStor v6.8.0.

Release Notes

Alerts and Configuration Analysis

  • Added alert webhook support for Google Chat. [ QSTOR-9793 ]
  • Allow users to have a subset of alert subscriptions. [ QSTOR-11683 ]

Ceph Scale-out Block, Object and File

  • Added an 'Enable Versioning' checkbox to the Create Bucket dialog (automatically enabled when object locking or governance/compliance is selected); versioning can also be enabled on existing buckets. [ QSTOR-10937 ]
  • Added Ceph Tentacle 20.2.2 support to the Jammy QuantaStor ISO and package mirror. [ QSTOR-11441 ]
  • QuantaStor 6.8 ships Ceph 20.2.2 (Tentacle), which includes the upstream MDS fix (tracker #76031) for the SIGABRT/crash on client eviction caused by a null MDRequestRef under sustained metadata contention. [ QSTOR-11504 ]
  • The Ceph Tentacle 20.2.2 release resolves frequent radosgw (RGW) crashes related to Kafka bucket notifications. [ QSTOR-11687 ]
  • Fix ceph bucket delete with large number of objects timing out. [ QSTOR-11728 ]
  • Removed: the deprecated S3 bucket orphan-object cleanup feature; any existing scheduled task is disabled automatically on upgrade. [ QSTOR-11729 ]
  • Fixed a regression where deleting multiple Ceph (RBD) volumes at once on a multi-node cluster would fail after the first volume. [ QSTOR-11756 ]
  • Added support for major-version Ceph upgrades (through Tentacle 20.2.x) in the upgrade script. [ QSTOR-11762 ]
  • Increase the timeout for "Add Node" to a ceph cluster [ QSTOR-11906 ]

Dashboards and Analytics

  • Added: a 'qs-statsdb resetpassword' command to rotate the InfluxDB stats password; telegraf is updated in lockstep so dashboards keep working. [ QSTOR-11170 ]
  • Rebuilt bundled telegraf with its vendored rclone library updated to 1.74.3, clearing CVE-2026-41176 and CVE-2026-41179. [ QSTOR-11771 ]

Enclosures and Controllers

  • Fixed issue with "disappearing" enclosures [ QSTOR-11676 ]
  • Added support for the Supermicro SSG-222B-NE3X24R all-NVMe server, including hardware enclosure auto-detection, enclosure layout and diagram, and NVMe drive slot mapping in the QuantaStor web management interface. [ QSTOR-11712 ]

High Availability Failover

  • Add logging for manually run iofence commands [ QSTOR-11751 ]
  • Fixed HA pool failover persisting a stale size=0 / isMounted=false when the pool's datasets were still mounting during fast-import; the last-known-good size and mount state are now preserved (and the corrected record wins via a bumped modified timestamp). [ QSTOR-11818 ]
  • Fixed: Storage Pool HA io-fencing and failover now work over NVMe-TCP; target namespaces advertise persistent reservations. [ QSTOR-11850 ]

Kernel and Drivers

  • Updated Jammy platform Kernel to 6.8.0-90-generic [ QSTOR-10903 ]
  • Updated Jammy platform ZFS driver to 2.3.5 [ QSTOR-10903 ]
  • Updated Jammy platform SCST driver to 3.10 [ QSTOR-10903 ]
  • Updated Jammy platform mpi3mr driver to 8.15.1.0.0 [ QSTOR-10903 ]
  • Updated the Mellanox OFED driver installer for current Ubuntu 22.04 (Jammy) kernels, fixing driver installation failures on recent kernels. [ QSTOR-11709 ]

Multitenancy

  • Added RBD namespace support, allowing multiple tenants to share a single Ceph block pool with isolated volumes. [ QSTOR-11650 ]

NAS Gateway / Cloud Containers

  • Add a fix for Samba access to Cloud Container shares. [ QSTOR-11763 ]
  • Fixed: cloud containers are now accessible over SMB, not just NFS. [ QSTOR-11824 ]
  • Updated the bundled AWS CLI to version 2.35.6. [ QSTOR-11827 ]

Network Interfaces

  • Added DOCA OFED install script for jammy and upcoming noble platforms as an optional install for latest hardware. [ QSTOR-11069 ]
  • Fixed the Network Interface Rename feature on systems upgraded from older releases where the 'ip' command resides in a different path. [ QSTOR-11614 ]
  • Fixed: converting a local VIF to an HA VIF now preserves the original local interface if the conversion fails, so client connectivity isn't lost. [ QSTOR-11706 ]

Network Shares

  • Add the ability to specify per user or group permissions on Network Shares during create or modify [ QSTOR-10662 ]
  • Fixed: intermittent 'NFS server not responding' stalls under load by raising the nfsd listen backlog from 64 to 4096. [ QSTOR-10872 ]
  • Fixed: NFS client entries can now be added or removed on a share while a long-running replication or snapshot task is using it, instead of queuing. [ QSTOR-11571 ]
  • Consolidated the Network Share create/modify dialog's Permission Settings into the File Masks tab — now a single 'File Masks & Permission Settings' tab containing Owner Permission Settings plus the Directory and File Create Permissions Masks. [ QSTOR-11625 ]
  • Reduced excessive event logging (event spam) for network shares and related grid objects. [ QSTOR-11695 ]
  • Fixed handling of CIFS share options so empty/null option values are no longer written to the Samba configuration. [ QSTOR-11731 ]
  • Fixed an issue where an Active Directory group added to a share's permissions was not preserved, and excluded built-in/default AD groups from the group picker. [ QSTOR-11767 ]

Remote Replication, Snapshots and Backup Policies

  • Copy source Network Share settings to the destination share after replication. [ QSTOR-5738 ]
  • Add support for rollback of Storage Volumes and Network Shares for a replication schedule [ QSTOR-6227 ]
  • Minor updates to the auto snapshot cleanup tasks for replication schedules. [ QSTOR-11622 ]
  • Fix auto create of replication association after failover [ QSTOR-11700 ]
  • Snapshot schedule now shows the state detail when schedule is disabled by user. [ QSTOR-11717 ]
  • The Snapshot Schedule properties page now displays Retention Tags. [ QSTOR-11719 ]
  • Changed: Auto-tier backup policies now display as 'Auto-Tier >> Outbound' instead of 'Move >> Outbound'. [ QSTOR-11764 ]
  • Fixed the CLI replication-schedule rollback so volume and share names are accepted (not just IDs), and rolling back only volumes (or only shares) no longer also rolls back the other type. [ QSTOR-11814 ]

SCSI / NVMe Target

  • Added: a 'Select Initiator...' picker in the Add Host dialog to choose a discovered FC remote initiator WWPN instead of typing it. [ QSTOR-10741 ]
  • Renamed the sys-list 'Firmware Version' column to 'SCSI Target Version' and corrected the value to show the actual SCST target-driver version (e.g. scst-3.10.0.qs01-...) instead of the literal 'scst-SCST'. [ QSTOR-11854 ]

Security

  • Added support for Okta multi-factor authentication. [ QSTOR-9399 ]
  • Added Multi-Admin Approval support, allowing deletion of pools, buckets, shares, and volumes to require approval from a configurable number of administrators. [ QSTOR-11344 ]
  • Fixed certificate update failures when using a certificate template, and ensured dependent services (such as HAProxy and RGW) are restarted after a certificate replacement. [ QSTOR-11678 ]
  • Fixed an issue in the Security Manager dialog where it was showing the "password policy change" warning confirmation when no fields were changed. [ QSTOR-11708 ]
  • InfluxDB (port 8086) now requires authentication and binds to localhost by default; remote access is opt-in via the tf_influxdb_remote_access.enable touchfile (authentication still enforced). [ QSTOR-11759 ]
  • Added the Ubuntu security apt repositories to the default package sources and removed unused bundled components flagged by scanning (Chronograf, the LXD/core20/snapd snaps, the docker-scan plugin), clearing CVE-2024-41110, CVE-2026-34177/-34178/-34179, CVE-2020-14343, CVE-2025-68121, and CVE-2024-45337. [ QSTOR-11760 ]
  • Rebuilt the bundled QuantaStor Go tools (qs_iofence, qs_pool_export_helper, qs_diskinfo) and s5cmd with a current Go toolchain, clearing CVE-2025-68121, CVE-2024-45337, CVE-2024-24790, CVE-2023-24540, and CVE-2023-24538. [ QSTOR-11761 ]
  • Hardened the internal QuantaStor script service (scriptsvc): removed a command-injection vector in request handling and now enforce the service's authorization check on every call. Addresses ZDI-CAN-30036 [ QSTOR-11799 ]
  • Made a number of usability and reliability improvements to the multi-admin delete-approval workflow. [ QSTOR-11804 ]

Service Core

  • Fix tag-set where custom key/value tags were not appearing for User objects. [ QSTOR-11528 ]
  • Improved Ceph OSD daemon restart handling during upgrades and reduced database load on large-scale clusters. [ QSTOR-11675 ]
  • Fixed an issue with QuantaStor management service memory usage when a grid node was offline. [ QSTOR-11693 ]
  • Improved crash handling and diagnostic tracing for coordinated tasks, during operations such as adding a Ceph RADOS Gateway across multiple nodes. [ QSTOR-11702 ]
  • Improved logging and error checking for QuantaStor configuration database backups, including capturing error output when a backup fails. [ QSTOR-11742 ]
  • Expanded diagnostic log collection to capture additional network routing details (ip route output and static-route configuration). [ QSTOR-11748 ]
  • Fixed the Sessions view showing stale SMB/NFS client sessions that lingered after a reboot; the share session monitor now forces a full re-sync on its first poll after startup so sessions left over from before the reboot are cleared. [ QSTOR-11757 ]
  • Fixed a self-sustaining grid event flood triggered when creating CephFS/CIFS network shares at scale: nodes no longer re-cache remote-origin (sync-applied) events, and CIFS options are diff-merged to secondaries, eliminating the event resync cascade. [ QSTOR-11785 ]
  • qs-sendlogs now writes a qs_logs.runtime.info file recording the options and variables used during log collection, and reliably scrubs sensitive data from the collected database. [ QSTOR-11897 ]

Storage Pool Management

  • Added automated ordered-disk-list pool creation for Seagate Corvault and WD Data24 systems, simplifying creation of fully redundant storage pools. [ QSTOR-11538 ]
  • Fixed software-encrypted pool creation so a disk block-size mismatch is detected before disks are encrypted, avoiding leftover encryption mappings. [ QSTOR-11599 ]
  • ZFS deduplication now uses the faster fast_dedup mode by default, improving sustained deduplication performance. [ QSTOR-11714 ]
  • When a LUKS-encrypted disk member of a pool momentarily goes offline and the hot-spare swap-in completes, if the "bad" disk in question has in the meantime come back online, its iofence status, encryption status and its related properties in its respective disk object are cleared up, making the disk available for use again. [ QSTOR-11738 ]
  • Resolved a boot-time race where the OS-level zfs-import-cache.service could import ZFS pools before qs_service started. QuantaStor now imports pools with cachefile=none, masks the zfs-import-cache and zfs-import-scan services on install and upgrade, and clears the stale global ZFS cache file at boot, ensuring QuantaStor remains the sole coordinator of pool imports. [ QSTOR-11752 ]
  • Add the ability to create a Storage Pool checkpoint for scale-up pools [ QSTOR-11784 ]
  • Added a support-only, touchfile-gated pool-export triage call-out (qs_pool_export_triage.sh) that gathers diagnostics when a share or volume is still open during a storage pool export. Off by default (enabled via the tf_pool_export_triage.enable touchfile). [ QSTOR-11807 ]
  • Resolved issue preventing HA failover functionality using SED drives. [ QSTOR-11832 ]

Upgrade Manager and Installer

  • Added retry logic with a timeout to the Ceph upgrade script's daemon restarts so transient restart failures no longer abort the upgrade. [ QSTOR-10598 ]
  • Fixed focal-to-jammy distribution upgrades leaving the HA cluster service (pcsd/pacemaker) inoperative due to a missing ruby-webrick dependency on Ruby 3.0+. The upgrade now installs ruby-webrick so HA failover keeps working. [ QSTOR-11841 ]
  • Fixed focal-to-jammy distribution upgrades that could abort and roll back when Ubuntu's unattended-upgrades/apt-daily held the apt lock or needrestart blocked on a prompt. QuantaStor now disables automatic apt upgrades and runs the distribution upgrade non-interactively. [ QSTOR-11874 ]
  • Fixed a Noble (24.04) upgrade failure where qs-upgrade aborted because the dpkg package version string could not be parsed. QuantaStor now compares Debian package versions using dpkg's native version ordering. [ QSTOR-11881 ]

Web Manager

  • Minor fix to the "Copy to Clipboard" for task descriptions [ QSTOR-9574 ]
  • Added a method field and a sortable timestamp column to the web interface Audit Log Analyzer. [ QSTOR-11191 ]
  • Updated web interface translations across the major supported languages. [ QSTOR-11585 ]
  • Added a 'Create Share' option to the share template menu to quickly create a share applying all of a template's settings. [ QSTOR-11595 ]
  • Fixed the web interface to display all configured Ceph MDS daemons instead of limiting the list to five. [ QSTOR-11615 ]
  • Add auto select convenience buttons to the Storage Volume multi-delete dialog. [ QSTOR-11707 ]


Security Notes

12 security notices published since v6.7.6.001 (June 25, 2026) affect this release: 3 fixed in this release, 9 fixed by a QuantaStor security update (3 externally-facing, 9 internally-facing). See QuantaStor Security Notices for v6.8.0.205.

Archived Change Logs

Change Log Archive 2026

Change Log Archive 2025

Change Log Archive 2024

Change Log Archive 2023

Change Log Archive 2022

Change Log Archive 2021

Change Log Archive 2020

Change Log Archive 2019

Change Log Archive 2018

Change Log Archive pre-2018

Select the link above to see the Change Log Archive of older revisions.