SoftNAS 5 Release Notes

• Additional Monit Checks - pool capacity checks.
• DriftBot - Removed from product.
• Lift and Shift - provide a way to cancel automatic share/volume size calculation.
• Nginx - Updated version to resolved outstanding CVEs.
• Red Hat Enterprise Linux 8.4 - support added in response to v8.2 end of support.
• Rocky Linux 8.5 - moved to Rocky Linux due to End of Support for CentOS Linux.
• Slack and Microsoft Teams - added as options for automatic monitoring notification.
• Snapshot scheduling - enhanced scheduling user interface and added a safe guard to ensure users do not delete volumes when trying to delete snapshots.
• Softnas-cmd - added iSCSI specific commands along with adding a flag to "executeupdate" to be able to only update using minor version.
• StorageCenter Administrative Interface - enhanced overall look and feel of the user interface.

• CIFS - resolved issue where a share was not being properly restored from a pool backup.
• DeltaSync - resolved an issue where opened Microsoft Office documents were not being properly handled.
• Install/Update - addressed several areas of cleanup regarding some minor errors reported in the software update details.
• Lift and Shift - improved performance and resolved an issue where new flows using a Private CIFS Share failed.
• NFS - increased NFS thread count and NFSv3 firewall port management corrections.
• SNAP HA/SnapReplicate - resolved an issue where status was not being properly displayed when communication was lost between the nodes, also resolved a calculation error that resulted in the progress of a replication to be reported incorrectly.
• StorageCenter Administrative Interface - resolved issue where the root disk of a VMware deployment was being displayed under Disk Devices, also resolved a couple of broken links within Quick Helps and Documentation.

• Install/Update - when performing an update using the StorageCenter Administrative Interface, the progress bar will seem to be "stuck" even though the update is still proceeding as expected.  If you are using this method to perform the update, expanding the Update Details pane will permit you to see the actual progress still being made. This is due to the Nginx update to this release. 

• Enhanced security by updating config file permissions.

• Disabled NFS ID mapping may impact ability to perform updates behind a firewall.

• Deployments using S3 or Azure Blob storage may not successfully update without configuring end points.

• Snapshot and Disk Activity may not reflect location time.

• VNET location may not be found when deploying new HA configuration.

• Downsizing of an AWS instance may result in boot failure.

• Samba may crash resulting in SoftNAS drives unmounting frequently.

• Nginx may not properly handle SSL certificates.

• SoftNAS NFS Thread Counts low due to nfsd not starting on new deployments.

• Identified “race condition” causing HA/SnapReplicate to fail manual takeover.

• Monit action blocking HA giveback.

Known Issues 5.1.3

Yum update - it has been observed occasionally when running a "yum update" from the command line you may experience an error message when the kernel debug package is installed.  This is contributed to the fact SoftNAS uses a custom kernel package.  The error can be ignored and will be resolved in a future release.

winbindd log message - There are some situations where RPC functions return RPC faults and this is not a fatal condition.  In that situation a dcesrv_call_dispatch_local: DCE/RPC fault in call lsarpc:32 - DCERPC_NCA_S_OP_RNG_ERROR message will be observed in the /var/log/messages file.

• Snapshot and Disc Activity may not reflect location time.

• VNET location may not be found when deploying new HA configuration.

• Downsizing of an AWS instance may result in boot failure.

• Samba may crash resulting in SoftNAS drives unmounting frequently.

• Nginx may not properly handle SSL certificates.

• SoftNAS NFS Thread Counts low due to nfsd not starting on new deployments.

• Identified “race condition” causing HA/SnapReplicate to fail manual takeover.

• Monit action blocking HA giveback.

• SnapHA prematurely starts before remote node is ready.

• Manually creating NFSv3 shares lack write privileges.

• Increased NFS thread count for improved performance.

• Security updates and fixes, notably CVE-2021-3177.

SmartTiers Pools/Volumes/SnapClones not remounted issue - An issue occurring in some SmartTiers deployments, where pools, volumes, and associated SnapClones were not mounting after a reboot has been resolved. 

HA installation failure issue - Resolved an issue where high availability (SNAP HA) would fail to install due to an assumption that the two VMs were in the same resource group. Logic added to ensure that subnet info is consulted prior to connection attempt. 

RHEL Root Corruption and Disk Unmountable issue - A known RHEL issue related to THP (transparent huge pages) default setting resulted in metadata expanding beyond memory capacity, resulting in data corruption. THP is now disabled at boot by default to prevent this issue re-occurring. 

Unable to create Azure Blob (object) Storage - An issue in which creation of blob storage on Azure would result in "unknown error" has been resolved. 

Invalid Help Link when connecting with SSH - Updated support link displayed when connecting by SSH to your SoftNAS instance/VM to ensure customers have easy access to support. 

Updated video links and content for in-app help - Updated or added video links to in-app help items, and added new help content to app features. 

Lift and Shift Azure Storage Type Selection issue - An issue in which the customer could not select disk type for Lift and Shift flows on Azure instances has been resolved. 

Failed takeover from UI - A race condition that could result from manual takeover/giveback operations that could result in takeover failure due to instance/VM roles switching places too quickly has been resolved. 

Failed error reporting issue - In some isolated cases, SnapReplication would not report failure if previous cycle completed successfully, and the next cycle was unable to start due to target node being down. This could result in replication showing as functional when it was not. This has now been resolved. 

Lift and Shift Home (Flexfiles) Repository available mappable target - An issue in which the home repository for Lift and Shift was listed as available as a target for mapping (Lift and Shift should not target the home directory) has been resolved.

Lift and Shift Failure to Create Repository - An issue in which attempts to create a Lift and Shift repository has been resolved. (19181)

SSL Certificate Issue between VMware and AWS Lift and Shift operations - Necessary communication between VMware and AWS instances for Lift and Shift operations were being blocked by certificate configurations. This has now been resolved. 

SnapReplicate Throttling issue - An issue where throttling command was not installed by default, resulting in potential poor performance, has now been resolved.

Nginx comment line issue - An issue where a SED REGEXP would constantly add # to the nginx registry file whether commented already or not, potentially resulting in large files sizes over time, has now been resolved. 

Nginx SSL certification corruption issue - An issue in which failover for HA would not work due to Nginx SSL Cert corruption has now been resolved. 

Custom NFS IDmapping configuration issue - An issue with a custom script in a client environment has been connected to an unsupported NFS setting. NFS IDmapping has been disabled in SoftNAS 5.1.1 for security reasons. Enabling NFS IDmapping is not a recommended configuration. 

• Added needed support for RHEL 8.2 and supporting installer.​

  • SoftNAS RHEL 8.2 BYOL Only​

  • Installed onto an existing fresh install of RHEL 8.2​

  • Installed via an installer script which will be downloaded from our website.  ​

The installer and documentation can be found at: RHEL SoftNAS Installation Steps

• Ability to enable FIPS​

• Added AWS SSM agent

Failover Issue when using custom SSL certifications - An issue in which custom SSL certifications would be over-written during a failover operation, resulting in loss of high availability, has been resolved. The custom SSL certifications will no longer be over-written. 

Error when modifying NFS shares from UI - An issue in which attempts to modify an NFS share from the Create an NFS Share section of the UI would result in failure and an error message have been resolved.

Security Fixes - The following fixes have been made to secure SoftNAS 5.1.0:

• removed vulnerability to CVE-2021-3156 (sudo related) 

• removed kernel vulnerabilities to CESA-2020:4286, CESA-2020:4431 and CESA-2020:4685 ​

• Netatalk completely removed from the system

Yum Update Issue - An issue preventing yum updates from completing has been resolved.

Benign Log Error - Log errors of a specific type seen occasionally on install has proven to be benign. The error reads  "Job for <service name> failed because the control process exited with error code"

Service names that have been seen for this error include:

• webmin.service

• php-fpm.service

• sernet-samba-smbd.service

• sernet-samba-winbindd.service 

These log errors will be addressed in a future release.