v4.4.0.174 (November 13th 2017)

Upgrade Instructions

Click here for instructions on upgrading to QuantaStor v4.4.0.

ISO/DVD/USB Boot Install Image

Release Notes

New Features

  • Added New Grid Dashboard tab to the Web Manager that provides a quick at a glance overview of Resource, Cluster, and System Health and Status for nodes in a QuantaStor Grid.
  • Added qs-upgrade script to provide Distribution Upgrade support to migrate QuantaStor Appliances from 12.04 Precise to 14.04 Trusty. Note: For Upgrading HA Clusters or Scale-Out Configurations, Please contact OSNEXUS support for assistance.

Active Directory

  • Fixed: User Access assignments will persist after Active Directory server configuration has been removed. Previously these settings would have been removed on leaving the Active Directory domain.
  • Fixed a rare case where response from Active Directory server is slow, the join domain task could fail.

Remote Replication and Snapshots

  • Added Features to Snapshot Schedules Backup Policies to bring them in line with those already available with Remote Replication Schedules. This includes new Interval based Timers and Long Term Retention Tagging and Policies.
  • Fixed a filtering item with Remote Replication and Snapshot schedules that could have limited the Volumes and Shares available for selection. Now only Volumes and Share on the destination pool for Remote Replication schedules is filtered out from the list of available replication sources.
  • Fixed an issue where nodes would sometimes report being out of sync during a remote replication.
  • Fixed: Increased the maximum retries and wait time for Remote Replication snapshot discovery to better support slower replication target systems.
  • Fixed: Corrected an issue where the Schedule ownership was not being set on a Remote Replication Association if the same association was used for Manual Remote replication as well as via a Remote Replication Schedule.
  • Fixed: The associated Replication Task will now fail as expected when a Replication processes between nodes is terminated due to network or system stability issue.
  • Fixed: The Remote Replication Task now provides better logic for the Information used by the Remote Replication Reports to show the start-times, end-times and replication speeds.
  • Fixed: Replica _chkpnt snapshots will now have their share parentId set correctly for their parent target _chkpnt of a remote replication.
  • Fixed: the Remote Replication task status will correctly update with a failed status if the replication process is terminated due to a communication or stability problem on the source and/or replication target systems.
  • Fixed: Create Remote Replica Volume Tasks will now correctly detect there is no common snapshot for delta's between the source and target and fail the task. Previously the task would stay at 0% and running status and never complete or fail.
  • Fixed: Replication Network Share snapshots on the source system could sometimes not be correctly associated with the Replication Schedule that created them. This is now fixed.
  • Fixed the descriptive and title text in the Snapshot Modify Dialog.

Ceph Scale-out Block and Object

  • Added new 'Enter/Exit Ceph Maintenance Mode' Dialog to allow Administrators to set the Maintenance Mode state on a Ceph Cluster.
  • Added 'qs ceph-osd-replace-journal' command to allow Administrators to replace a Journal device in an OSD live. Note; The ceph cluster must be in maintenance mode to allow this option and a restart of the OSD receiving the journal change will occur.
  • Added 'qs enter-system-maintenance' and 'qs exit-system-maintenance' CLI commands. This is currently implemented only to set Ceph Scale-out Block and Object clusters into maintenance mode but will be expanded to support maintenance mode management of other QuantaStor Cluster/scale-out solutions.
  • Fixed an issue where a Ceph Cluster Node that had RBD devices mapped via iSCSI could not be removed from the Ceph Cluster.
  • Fixed: The Ceph Cluster Dashboard will now correctly display if you select a Journal Device that is not assigned to an OSD.
  • Fixed: Selecting an object in a grid view in the Ceph Scale Out Block and Object section of the Web Manager will now select the parent object in the tree.

Storage Pool

  • Added a larger 4GB zfs_dirty_data_max setting for systems with 16GB of RAM or more. For systems with less than 16GB of RAM, the default 1GB cache setting will be used.
  • Fixed: Create Storage Pool Tasks on Encrypted disks now provides a more detailed Task description while at the Encrypt Disks stage.
  • Added a status column to 'qs pool-list' to show the Storage Pools reported Health Status.
  • Added an alert recommending a change of sync policy to 'standard' If a ZIL configuration is removed from a ZFS Storage Pool that has had sync=always configured. This is to ensure an expected level of performance can be maintained after removal of the high IOPS ZIL Slog(Sync Log) SSDs. A policy of sync=standard is generally recommended for all ZFS configurations unless advised by OSNEXUS Support due to the needs of a specific use case and workload.
  • Fixed an issue where a spare device could be rediscovered and added back in after removal from an XFS pool.
  • Fixed: Alerts will correctly be triggered when the Storage Pool Used Capacity Thresholds are Exceeded.
  • Fixed: XFS Storage Pools will correctly show the add/remove hotspare context menu in the Storage Pool section for adding dedicated hot spares to XFS Software RAIDed pools.
  • Fixed: The Color Thresholds for Utilized % in the Storage Pool section of the Web Manager is now based off of the values defined in the Alert Manager Pool free space thresholds.
  • Fixed an issue where the RAID levels drop down in the Storage Pool Create Dialog would not always update to reflect the specific options available for the ZFS or XFS filesystem type chosen. This also corrects an issue where the RAID levels listed would not always uopdate based on the number of physical disks in a selected system in the dialog.

Networking

  • Added an additional LACP bonding mode that enables support for Layer 3+4 xmit_hash_policy. The original LACP mode is still identified as lacp via the CLI and will show 'LACP Layer 2' in the Web Manager.

Disk Management

  • Added: The Format Physical Disk Dialog now provides the ability to select multiple disks. The Disks listed are from the detected available disks on a Storage System that are not associated with a Storage Pool.
  • Added: The Hardware Disk Column is now visible by default for all dialogs that interact with Physical disk Objects.
  • Added a Search field to the Physical Disk Tree View. The disks can be searched and filtered based on Disk logical Name and Serial Number.
  • Added: Newly inserted Hardware disk devices will automatically trigger a Physical Disk Scan and be immediately ready for use.
  • Fixed: increases fs.aio-max-nr value to 1048576 to support larger initial multipath configurations by default.
  • Fixed: Added Paging to the central grid pan in the Physical Disk View Section of the WebUI.
  • Fixed: added further performance optimizations to the Physical Disk Section of the Web Manager.
  • Fixed: Further Enhancements to the Format Disk functionality.
  • Fixed: Text in the Physical Disk Grid view can be selected allowing for easier copy/paste of information.
  • Fixed: the SerialNo. Field is now visible by default in the Grid view.

Hardware Enclosures and Controllers

  • Added SMART health status in the Hardware Controllers and Enclosures section for disks connected to SAS HBA's.
  • Added: The Hardware Disk Identify Dialog in the Hardware Enclosures and Controllers section of the Web Manager now supports selecting multiple Hardware Disks at the same time.
  • Added: the Hardware Enclosure view in the Hardware Enclosures and Controllers section of the Web Manager now will display the pool the disk is associated with.
  • Added: The Mark/Unmark Hotspare Dialog in the Hardware Enclosures and Controllers section of the WebUI Now presents a dialog capable of performing multiple selections at once.
  • Fixed an issue where LSI 9400 series HBA's would not show disk temperatures or trigger overtemp alerts.
  • Fixed: The Enclosure grid view in the center pane will now update as expected based on the selected Hardware Controller in the Tree view.
  • Added support for newest HPE RAID controller management via ssacli 3.3.10.3.0
  • Fixed: Changed the storcli discovery support for SAS HBA's to only be enabled for LSI/Broadcom Branded 9400 series controllers.
  • Added further Search filtering criteria (Vendor, Product, Controller, Status) for the Hardware Disk Identify and Mark Hot Spare Dialogs.
  • Added: Hardware Disk Identify Dialog is now available as a right click context menu item for controller, enclosure and other items in the tree view of the Hardware Enclosures and Controllers section of the Web Manager.
  • Added: Hardware Disk Identify now supports 'On' and 'Off' modes in addition to the existing Duration mode previously offered. Now you can set a Disk to Indicate it's location without having to race the clock.
  • Fixed an exception where the Hardware Controller->Create Raid Unit, Software Controller->Remove Adapter, and Software Controller->Scan Targets dialogs would not open if there were no available controller objects to perform actions on.
  • Fixed: added logic to handle a corner case where a failing or Hardware RAID controller in an errored state would return a status of 'Failed' via it's raid utilities but still provide valid parsing data on the state of Enclosures, Disks and RAID units behind the controller. Previously QuantaStor trusted the 'Failed' status code returned from the card and stopped discovery at that stage.


Security

  • Added NIST 800-53 R4 AC2(2) compliance to support emergency and temporary account type flags for users created via the user-add qs CLI command. Please refer to the qs user-add command help '--account-type' and '--hours-term' flags for further details.
  • Additional NIST AU-12 Compliance
    • QuantaStor v4.4 now outputs to the qs_audit.log file in CEE JSON reference profile conformant to RFC 4627.
  • Added: QuantaStor 4.4.0 now deploys SSL certificates on new installs with sha256 signed 2048-bit SSL Keys. It is recommended that you upgrade all older deployments to use the newer SSL certificates using the 'qs-util cacertusedefault' command from the CLI. If you have a deployment that needs to continue to use the older SSL Certificates, they are still available by running 'qs-util cacertuselegacy' All QuantaStor grid systems must be running the same type of signed keys for grid communication to function.
  • Added: System Swap space is now automatically encrypted when an Encrypted Storage Pool is created or detected on a system after updating to the 4.4.0 release.

Web Manager

  • Added: Tree View Search Filter now returns results where child objects matched.
  • Added Further Localization enhancements to Dialog text.
  • Fixed an issue with some dialogs containing lots of columns, where you could not always scroll fully to the right.
  • Fixed: The Dialogs involving Physical Disk and Hardware Disk objects now show a disk total at the bottom of the dialog that reflects the number of disks found from the current search filter. If there is no search filter, it will show the total number of disks available in the system.
  • Fixed an issue where the Utilized % Column in the grid view would not show 100% Utilized when a Pool was full.
  • Fixed an issue where the Utilized % Column would not update as expected for Storage Volumes after a discovery cycle.
  • Fixed: The Add Grid Management Virtual IP Dialog now has the correct help web page link specific to creating a Grid Management Virtual IP.
  • Fixed: the right Click Context Menu for Gluster Volumes has been re-ordered to ensure a clearer order of possible operations.
  • Fixed: Updated the Workflow Manager help links and images in the Documentation wiki.

Service Core

  • Updated SNMP MIB
  • Fixed: Changes to the Password and Security Policies no longer fail if a node does not yet have a active QuantaSor License.
  • Fixed an issue where Task description text was not updating as intended for running tasks.
  • Fixed an issue where the sync policy setting on ZFS Network Shares was not being set if a user wanted something other than default inherited from the ZFS Storage Pools setting.
  • Fixed: Improved logging to qs_service.log for Active Directory join.
  • Added ZFS Event Daemon(ZED) to rsyslog for event logging.
  • Fixed an issue where the ZFS Event Daemon(ZED) was not starting on startup. ZED will now start and be kept alive by the normal QuantaStor keep alive scripts.