v4.3.0.335 (June 22nd 2017)

Upgrade Instructions

Click here for instructions on upgrading to QuantaStor v4.3.0.

ISO/DVD/USB Boot Install Image

 

Release Notes

Ceph Scale-out Block and Object

  • Updated Ceph Packages available:
- 10.2.7 Jewel for Trusty based installs 
- 0.94.10 Hammer for Precise based installs
  • Fixed Ceph Scale-out Block startup discovery on grid nodes that are still part of a ceph cluster, when one of the ceph nodes in the grid is removed from the grid.
  • Fixed: Added validation to correct an issue where a Grid Node Object and associated child objects were unexpectedly removed from the Grid Master if the QuantaStor Grid Node came onto the network with a new Storage System ID but using the same IP. This corrects a scenario where a system reinstalled in place due to a Hardware issue could cause an unexpected grid or configuration change.

 

Cloud Containers

  • New IBM Softlayer S3 Endpoints added to the default Cloud Providers configuration.
  • Fixed an issue where changes to the /etc/qs_cloud_providers.conf on the master node was not propagating to secondary nodes.
  • Fixed an issue where the entry fields for Tag and End-Point in the Cloud Containers> Add Provider Location Dialog and CLI were swapped.
  • Fixed a missing python-six dependency issue with the aws CLI tool.

 

High Availability

  • Improved pool failover times when network interface connectivity between nodes is lost.
  • Fixed an issue where high-availability virtual network interfaces could try and start on a system that did not currently have the QuantaStor service running.
  • Fixed issue where deleting high-availability virtual network interface could cause temporary outages on other high-availability virtual network interfaces.
  • Fixed an issue where deleting a high-availability virtual network interface could cause an unnecessary failover to occur.
  • Fixed an issue where failover would fail due to not finding the disks correctly on the secondary node.
  • Fixed an issue with nginx web service starting during qstormanager package install.
  • Fixed: HA Storage Pool Failover is now kernel panic aware to trigger a failover to the secondary node.
  • Fixed: Deleting a HA VIF from the Network port list in the Storage System View now correctly cleans up the High Availability Failover Group HA VIF object.
  • Fixed: A Site Cluster can now be deleted using the force flag if a Cluster node is Permanently offline and will not be returning.

 

Disk Management

  • Added a Multipath Configuration Dialog to the Physical Disk section of the Web Manager, this allows for administrators to scan for SAS, FC, iSCSI and other multipath/multiport capable devices and add their white-listing rules to the multipath configuration. This functionality is also available via the new qs CLI 'disk-multipath-config-list, disk-multipath-config-scan, disk-multipath-config-add and disk-multipath-config-remove' commands.
  • Fixed an issue where the /dev/disk/by-id/ata-* devices would be removed under the device mapper path after a Storage Pool is deleted by a user. Previously, a udevadm-trigger command would be required to bring the ata-* device back.
  • Fixed an issue with physical disk scan on multipath configurations that could cause the mutlipath devices to not appear in the QuantaStor WebUi and CLI once the scan completes.
  • Fixed: the qs disk-scan command now has a force flag


Storage Pool

  • Fixed an issue where old disk encryption keys would not be cleaned up after a storage pool is deleted.
  • Added earlier validation checks for Storage Pool Grow Operations for Encrypted Storage Pool configurations to ensure that the disks to be added are not encrypted before the RAID set size and resultant configuration is confirmed to be valid.
  • Fixed an issue during storage pool create where user specified RAID set sizes would not be used.
  • Fixed a few scenarios where creating a Storage Pool with XFS, Software RAID and Encryption could fail.
  • Fixed an issue where a failed disk device was not automatically removed from a ZFS Storage Pool in Multipath configured environments.
  • Fixed an issue that could occur when adding hot spares to ZFS Storage Pools in Multipath configured environments.
  • Fixed an issue where a failed disk device was not automatically removed from a ZFS Storage Pool in Multipath configured environments.

Remote Replication and Snapshots

  • Added a Remote Replication Report tab to the Remote Replication Schedules section of the Web Manager that shows the results of past replication tasks. Statistics in the report include: completion status, average throughput, the start and end time of the replication and many more. This is also available via the qs CLI with the 'qs replica-report-summary-list' and 'qs replica-report-entry-list' commands.
  • Added new snapshot retention options to the Create Remote Replication Schedule Dialog to allow for Daily, Weekly, Monthly and Quarterly Snapshots for Historical Storage Volume and Network Share snapshots.
  • Added new snapshot tags for Daily, Weekly, Monthly and Quarterly Snapshots that correspond with the retention policy picked for a particular Storage Volume or Network Share snapshot.
  • Added Compression support to Storage System Links, this allows for improved performance over slow WAN links.
  • Added: Remote Replication bandwidth throttling has been moved to the Storage System Link object. The qs link-create and link-modify commands and Web Manager Create Storage System Link and Modify Storage System Link Dialogs now allow for setting the bandwidth throttling.
  • Added the ability to turn on Unencrypted support for Remote Replication Storage System Links. This uses mBuffer to provide a high performance unencrypted channel for Remote Replication between QuantaStor nodes.
  • Added the ability to configure the Bandwidth Limiter in the Create and Modify Storage System Link Dialogs.
  • Added more information to indicate replication schedule health/state and cause of failures due to any miss-configuration or networking communication error.
  • Added: Snapshot Schedules can now be created for snapshots (allows Snapshot of Snapshot Scheduling)
  • Fixed an issue where remote replication could fail for manually initated replications.
  • Fixed: Max Replicas in Replication Schedules are now referred to more correctly as Max Delta Points, this clarifies more precisely how many Intermittent and hourly scheduled replica snapshot points are retained between a source and target replication association.
  • Fixed: The Remote Replication Offset interval for Hourly/Daily Replication in the Schedule Interval tab of the Create and Modify Replication Schedule dialogs now defaults to 0 minutes and can go to a max of 59.
  • Fixed: ZFS snapshots will once more be correctly removed upon the deletion of a Network Share snapshot. This corrects a regression from the 4.2.2 release.
  • Fixed: resolved an issue with updating the Timestamps for running remote replication tasks that could result in the remote replication link having incorrect progress information.

 

Gluster Scale-out File

  • Fixed an issue that could cause unexpected behavior with Gluster Peer and Volume objects showing correctly when Gluster is deployed n the same QuantaStor management grid as non Gluster nodes.
  • Fixed an issue that would prevent cleanup of Gluster Volumes in a configuration where the gluster bricks or underlying pools had been already removed.
  • Fixed: Re-ordered the Ribbon bar icons in the Scale-out File Storage tab.


Network Share:

  • Added the ability to see SMB session information for Network Shares under the new Web Manager Network Share>SMB Sessions tab and the 'qs share-session-list' and 'qs share-get' CLI commands.
  • Added the ability for users to create a snapshot of an existing Network Share snapshot (snapshot of a snapshot). This support is limited to custom named snapshots and not snapshots created by a schedule that has @GMT in the name.
  • Fixed an issue on Precise platforms with Network Share Snapshots for Windows SMB Shadow Copy and File Versioning support.
  • Fixed an issue where removing NFS access from a Network Share would collapse the tree to the first level in the Network Share tree view. Now the tree will stay as expected when a NFS Access object is removed from the share.
  • Fixed an issue in the Namespace Add/remove Network Shares Dialog where changing the namespace in the drop down would not always update the available selections.
  • Fixed an issue with the search filtering in the Add/Remove Network Shares to/from Namespaces.
  • Fixed an Issue with creating shares that have multiple '$' in the name.
  • Fixed an issue with snapshot mount directory cleanup after a network share snapshot has been deleted. Note, this prevents the issue from occurring going forward, If users ran into this case previous to upgrading to the 4.3 release they may need to manually remove old @GMT snapshot mount directories from the network share _snaps directory.

Web Manager

  • Added a new Password Policy dialog available under the Users and Groups tab in the Web Manager that allows Administrators to enforce password Requirements. This includes:
- Minimum password character length 
- Password expiration (in days). 
- Number of allowed login attempts 
- Minimum days to wait before password change is allowed. 
- Number of unique passwords before reusing a password is allowed.
  • Added: the Storage Volume Close Session dialog now shows a select-able list of the current sessions, clicking okay will close all the selected sessions on the SCSI Target.
  • Added: Network Share and Storage Volume Multi-Delete from the web manager can be used to delete share/volume and its child snapshots (select Delete Child Snapshots).
  • Added: The Network Shares and Storage Volumes Multi-Delete dialog now has the option to "Hide Snapshots", making it easier to select the parent share/volume to be deleted.
  • Added Passthru Storage Volume support to the Web Manager, Passthru Volumes can be created by right clicking on a Physical Disk and choosing the 'Create Passthru Volume' option.
  • Added new columns to the Create Storage Pool Dialog to show the Source System and Source Storage Volume to better allow customers to pick specific Passthru physical disks using QuantaStor appliances as backend storage for front end QuantaStor appliances.
  • Fixed: General responsiveness and UX performance improvements for the Web Manager on larger scale configurations.
  • Fixed: Provided a clearer message in the Create Pool dialog for when users do not provide matching passwords for the 'Encrypt Storage pool with Passphrase' fields.
  • Fixed: the Dashboards will enforce the use of https for their rest calls when the Web Manager is using https.
  • Updated the create and Modify Remote Replication Schedule Dialogs for a better workflow.
  • Fixed a problem under the Remote Replication tab in the Web Manager that could lead to a slow unresponsive Web Interface.
  • Fixed an issue where the Replication Targets tab under the Storage System Links could show as empty.
  • Fixed an issue where a newly created Host Group to not show the selected hosts in that group. Previously, a browser refresh was required to show them once more.
  • Fixed an issue with the Web Manager that would cause object status or task objects to not update or show as completed when a large number of events were received. Previously, a refresh of the Web Browser would have been required if this occurred.
  • In the Migration Edition Workflow Manager, the View Network Shares has been replaced with View SMB Connections.
  • Added a Task list counter to the Task list at the lower part of the Web Manager.
  • Added Block Size column option to the Storage Volume grid view.
  • Fixed: Removed the unsupported cloning options in the context menu for Network Share Alias or Subshare. Cloning should only occur at the Parent Share level.
  • Fixed an issue in the Physical Disk view that could cause the Firefox and IE Web browsers to report an unresponsive script warning.
  • Fixed a few areas in the Web Manager where large numbers of objects or events coming in could resulting in an unresponsive script warning for some Web Browsers.
  • Fixed an issue in the Storage Volume tree view where scrolling down and selecting a volume could cause the tree view to 'jump' up to the top of the list.
  • Fixed an issue with the tree view for Network shares that could sometimes show the NFS client access out of order with the associated network share snapshot.
  • Fixed an issue with truncation of some of the options in the Create and Modify User dialogs.
  • Fixed field, text, scroll and other alignment issues in various dialogs.
  • Fixed miscellaneous spellings in various Dialogs.
  • Fixed the Add user dialog descriptive text.
  • Fixed: Added a check to the Storage Volume Advanced Settings CHAP Username/Password to ensure that both Username and Password are supplied before clicking on OK.
  • Fixed: Corrected an issue under the Remote Replication Schedule View in the Web Manager where some items under the left hand tree view could not be selected if the same Network Share or Volume was also in under another schedule.
  • Fixed: Re-ordered the Ribbon bar icons in the Scale-out Block and Object Storage tab.
  • Fixed: Re-ordered the Ribbon bar icons in the Storage System tab.
  • Fixed: References for CIFS protocol in the Web Manager have been renamed or further clarified to SMB.
  • Fixed: The CIFS Configuration Dialog has been renamed to Active Directory Configuration.
  • Updated Web Manager splash to detail how to properly refresh the QuantaStor Web Manager on OS X.
  • Improved Web Manager responsiveness and performance.

 

Hardware RAID Support

  • Creating a Hardware RAID unit under Hardware Enclosures and Controllers will now create the HwUnit object for display immediately with discovery of properties occurring in the background. This provides better User Interface feedback when creating a large number of Hardware RAID units.
  • Fixed: added a guard to the Hardware RAID Controller SSD Cache Unit delete to prevent removal of the SSD cache if it is actively in use with Any other Hardware RAID units on that same RAID Controller.


Installer and Packaging

  • Fixed an issue with the .iso install media that required internet access for a install to finish.
  • Fixed: The Installer will now show eth* devices for UEFI BIOS installs
  • Fixed a problem with the latest qstortarget package compatibility with the older 3.19.0-29-quantastor kernel.


Localization

  • Updated Japanese, Chinese, French, Spanish, and Italian localizations for the QuantaStor Web Manager.

Security

  • Added: Users who are inactive for 30 minutes are automatically logged out of the Web Manager.
  • Added: Auto Logout clears all state information from the Web Browser.
  • Added CJIS Section 5.5.4 Compliance:
- System Use Notification is available with System Usage Notification Message field under Password Policy Dialog. 
  • Added CJIS Section 5.5.5 Compliance:
- Session Lock is available with Auto Logout value under Password Policy Dialog.
  • Added CJIS Section 5.4.1.1 Compliance.
CJIS Section 5.4.1.1 Events logging in the /var/log/qs_audit.log for: 
- Successful and unsuccessful system log-on attempts. 
- Successful and unsuccessful attempts to use access, create, write, delete or change permission on a user account or other system resource. 
- Successful and unsuccessful attempts to change account passwords. 
- Successful and unsuccessful actions by privileged accounts.
  • Added CJIS Section 5.6.2.1.1 Compliance.
Note: CJIS default password requirements compliance can be enabled under the Password Policy dialog in the Users and Groups tab. In the dialog, select Suggested Defaults and change password complexity to strong.
Detailed CJIS Section 5.6.2.1.1 Compliance, Password Shall:
- Be a minimum length of eight (8) characters on all systems. (Compliant & Enforced)
- Not be a dictionary word or proper name. (Compliant & Enforced, since QS v4.1.1)
- Not be the same as the Userid. (Compliant & Enforced)
- Expire within a maximum of 90 calendar days. (Compliant & Enforced, since QS v4.1.1)
- Not be identical to the previous ten (10) passwords. (Compliant & Enforced, since QS v4.1.1)
- Not be transmitted in the clear outside the secure location. (Compliant & Enforced)
- Not be displayed when entered. (Compliant & Enforced)
- Erase cached information when a UI session in terminated 
  • Fixed: Users created with the Cloud Admin and Cloud User Role can now change their own passwords.
  • Updated Samba packages are available to address CVE-2017-7494, please upgrade your system using the qs-upgrade CLI command to install these packages and bring your system current with all other security and stability fixes available on the package repository. Note: for customers who installed the sernet-samba4 packages on the older precise platform using the samba4-install script, a workaround to address this security alert is detailed in the KB article here: KBLINK
  • Added Logic to terminate SMB sessions for users who have had their user access removed.


Active Directory

  • Added support for RFC2307 configuration to the Configure Active Directory Dialog in the Web Manager for Trusty platforms or Precise installs that have the optional sernet-samba4 packages installed.
  • Added Trusted Domain checkbox to the Configure Active Directory Dialog box for Trusty platforms or Precise installs that have the optional sernet-samba4 packages installed. This enables trusted domain support for the CIFS/SMB service
  • Fixed: The Sernet Samba4 winbindd service will now be automatically started if it is detected that it is not running. This brings the Sernet samba service management inline with the standard Ubuntu Precise and Trusty Samba services.


CLI

  • qs CLI commands 'system-shutdown', 'system-restart' and 'system-upgrade' commands can accept a '--sys-list' argument with a comma delimmeted list of storage systems to perform Shutdown, restart or upgrade tasks on multiple gris nodes at once.
  • Fixed: 'qs set-tag' now allows the use of object UUIDs to set tags. Prior to this fix, only names were allowed to set tags.
  • Fixed: replaced the Parent Share ID with the human readable the Parent Share Name to the 'qs share-list' command for snapshots.
  • Added clarity to the help for qs commands such as 'pool-create', 'pool-grow' and others that have the "--disk-list" argument.
  • Updated the Help text and error responses for qs pool-create.
  • Fixed qs CLI cluster-ring-member-get command to provide better context with the --cluster-ring-member argument.
  • Added: Network Shares and Storage Volumes can now be deleted from the qs CLI using the flag '--delete-child-snaps'. Adding this flag will delete all the child snapshots. If snapshots are used by schedules then the additional '--flags=force' option should be used.


Core Service

  • Added a new QuantaStor Log collection tool with the below new features:
- adds support for uploading via https. 
- the tool will fetch an updated json definitions file if available from the OSNEXUS update servers before gathering logging data. This allows up to date fetching of diagnostics when working with the OSNEXUS support team. 
- The Send Log Report task will now show more detailed status on the log gather scripts progress. 
  • Fixed an issue where DNS entries added in the Web Manager Storage System Modify Dialog were not being reflected in the /etc/resolv.conf nameserver settings file.
  • Fixed an issue with the Web Manager Send Log Report task where the log would fail to upload but the task would return success. The task will now return as failed if the upload or gathering of the logs fails for any reason.
  • Fixed an issue that was blocking https access when the 'qs-util disablehttp' tool was used to turn off http port 80.
  • Updated SNMP MIB.