Resolved Issues in 5.1.0-SP40

ORION-191880: Resolved an issue that prevented showing the last upgraded CNode in the progress message produced during an upgrade procedure initiated via VAST CLI with the --cnodes-batch-size-percentage option specified.

ORION-190369: Enhanced upgrade routines to prevent a scenario where after completing a frozen upgrade, the NDU portion of the process could be started due to an INVALID_TARGET_VERSION error.

ORION-180395: Updated upgrade routines to prevent removal of the known_hosts file during a DNode OS upgrade.

ORION-187388: Resolved an issue that could cause the dbox add command to fail with the cannot create directory ‘/vast/backup’: Permission denied error.

ORION-194007: Resolved an issue which could cause the cluster to encounter an assertion failed: (eres != EStoreRes::OK) ranges_to_mark not empty!  error followed by denylisting @handle=<...> as traversal_type=19 resulted with res=PERSISTENT_ERROR and then locally denylisting shard_type=0 shard_id=<...> as op=DELETE_ELEMENT resulted with res=28 deny list alerts.

ORION-190969: Made updates to improve handling of deletions of an extremely large amount of directories with a lot of subdirectories and files in them while the cluster has VAST Catalog enabled.

ORION-190443: Optimized the mechanism of bulk permission updates to speed up handling of very large directories.

ORION-190034: Enhanced VAST Catalog transaction management to prevent a scenario that could cause automatic denylisting of ESTORE BIG_CATALOG operations because of a write-write conflict.

ORION-188621: Fine-tuned the mechanism responsible for handling massive deletions to resolve an issue where deletion of a very large amount of data caused the VMS to report nearly running out of metadata capacity.

ORION-186723: Resolved an issue that caused an automatic deny list alert due to the TRUMAN_WRAPAROUND_ERROR.

ORION-190107: Updated RDMA routines to avoid a flow where an attempt to create an NFS mount using RDMA over IPv6 could be rejected with the Connection refused error.

ORION-184380: Added a user-friendly error message to be reported in case an attempt to create an S3 bucket view fails with a permissions denied error.

ORION-206735: Made improvements to avoid VAST Catalog performance degradation on a cluster where prior to the upgrade, logging of protocol operations was done after enabling the feature to save protocol audit results to a VAST Database table (Settings -> Auditing -> General tab -> the Save audit logs to VAST DB toggle).

Note

Performance of queries against user-defined VAST Database tables may still be impacted after the upgrade.

ORION-189280: Made updates to prevent deletion of VAST Database elements even if the S3 lifecycle rules configured on the cluster stipulate the deletion.

ORION-189222: Resolved an issue that could case a CNode container to restart with the assertion failed: (_type_and_coding.type_info.nullable || (validity_bitmap == nullptr)) error when using VAST Database SDK to import data to a VAST database.

ORION-190681: Updated the logic used to cache user ID mappings (SIDs, VAIDs) to resolve an issue where a protected path could not catch up with its RPO after a NATIVE_REPLICATION deny list was cleared.

ORION-181089: Updated LDAP caching mechanism to prevent a scenario where after a VMS HA event, it was not possible to log in to the VMS using LDAP until the LDAP cache expired.

ORION-202338: Resolved an issue where the cluster did not raise an alarm when a DNode link went down and remained disconnected for a period of time long enough to avoid false alarming.

ORION-200202: Updated password validations so that an attempt to set a password that contains special characters for the vastdata user does not cause the No special characters should be in the password error.

ORION-194862: Updated view path validations to allow for view paths or SMB share names that include any UTF8 characters. Prior to this change, an attempt to create a view with a path that included a non-ASCII character resulted in an error.

ORION-191852: Improved the Data Flow polling to avoid creating a large number of SSH connections from the VMS container to localhost when Data Flow hostname polling is enabled on the cluster (Settings -> Dataflow Settings).

ORION-188852: Enhanced calculation of used capacity percentage so that it does not cause a bigint out of range errors when trying to view quotas via VAST CLI or VAST REST API.

ORION-185215: Updated handling of preferred CNode IDs so that an attempt to run the vms modfy --preferred-cnode-ids command does not fail with the cnodes_set_vms_preferred failed: CNodesSetPreferredResultCode.NOT_FOUND error.

ORION-180970: Resolved an issue where VMS occasionally could not return identity policies for a user group queried via VAST REST API (/api/latest/groups/?groupname=<...>&context=aggregated&amp;tenant_id=<...>).

ORION-175599: Resolved an issue where a newly created QoS policy could not be seen in VAST Web UI, VAST CLI or VAST REST API although it existed in the VAST internal database.

ORION-171871: Resolved an issue that could cause the VMS Preferred must not be True if there are less than 3 CNodes error to appear on attempts to enable the VMS Preferred option for a virtual IP pool that already existed and included more than three CNodes.

ORION-113520: Resolved an issue where one of DNode ports was falsely reported as faulty during a periodic pre-upgrade check.

ORION-194719: Updated the logic used to select tenants when creating a virtual IP pool in VAST Web UI so that the All tenants option creates a virtual IP pool for all tenants but not for the default tenant only.

ORION-191852: Updated Data Flow settings (Settings -> Dataflow Settings) to show the hostname polling interval that is currently used by the VMS.

ORION-176969: Updated the Settings -> Notifications navigation menu item to point to the correct notification settings page. Prior to this change, it opened the VMS settings page.

ORION-176945: Updated VAST Prometheus Exporter to avoid showing the text (MB/s) in metrics that refer to IOPS values.

ORION-175524: Made updates to avoid showing UNKNOWN port location for some ports listed in the Infrastructure -> NICs page.

ORION-184206: Updated the logic behind the cbox list and cbox show commands so that the commands do not fail with the Invalid value for `subsystem`, must be a value greater than or equal to `1` error if the cluster uses subsystem 0.

ORION-202803: Resolved an issue that could case a DNode container to restart with the spinlock lock takes too long: 1835, cannot kill locker silo error.

ORION-199929: Made updates to prevent a flow that could cause node containers running CentOS-based VAST OS to restart with the spinlock lock takes too long error during a switch firmware upgrade.

ORION-199409: Resolved an issue where the leader node could not connect to the DNodes due to the Failed to read the CC register SPDK error.

ORION-197587: Resolved an issue where the cluster encountered multiple high VMS memory usage alerts.

ORION-195783: Eliminated a flow that could, in some cases, cause multiple CNode container restarts accompanied by spinlock lock takes too long alerts raised for the DNodes if one of the CNodes has its link state constantly changing from up to down and back again.

ORION-193977: Resolved an issue where a CNode container restarted due to no fibers for incoming request errors.

ORION-193511: Eliminated a flow where the DTray reboot mechanism could get stuck, causing inability to restart the DTray and temporary service disruption.

ORION-189215: Resolved an issue that could cause high memory usage alerts on many of the cluster’s CNodes.

ORION-186568: Resolved an issue where after encountering link issues, the CNode container restarted due to the failed removing IP <IP address> from interface ens1f1 error.

ORION-186081: Resolved an issue where a failure of a DNode with Kioxia drives could lead to CNode container restarts and temporary service disruption.

ORION-186003: Resolved an issue that could cause a CNode container to restart with the Invalid permissions for mapped object error.

ORION-184360: Eliminated a potential race condition that could cause a CNode container restart with the assertion failed: ((_reentrancy_level) > (0)) (0 > 0) error.

ORION-184028: Updated failover mechanisms to eliminate a flow that could cause multiple CNode and DNode container restarts followed by a short service disruption after one of the cluster’s switches was rebooted.

ORION-183745: Improved handling of CNode and DNode communication issues to prevent a scenario where following an IB switch reboot, all CNode containers restarted and DNode failures were encountered with the can't deactivate dnode: can reach DU error.

ORION-181080: Enhanced handling of deletions to help avoid scenarios that could result in md_usage_state changed from ABUNDANT to SCARCE alerts.

ORION-180071: Resolved an issue that could cause periodic CNode container restarts with the failed to allocate vmsg args for GetUpdatedS3UsersParams error.

ORION-179428: Fine-tuned the mechanism of sorting stripes during defragmentation to resolve an issue that caused multiple the stripe is stuck alerts on the cluster.

ORION-174205: Updated cluster’s internal database query caching to eliminate an issue that could cause a CNode container to restart with the Address not mapped to object error.

ORION-159518: Resolved an issue that could cause a CNode container to restart with the timeout expired for life_type=0,life_gen=<...> (INGEST_READ) with 1 active jobs - timeout is 300 seconds, and diff is 301 seconds error.

ORION-190436: Resolved an issue where multiple DNode containers restarted during a short period of time due to list_del kernel errors.

ORION-173555: Updated the logic behind CNode selection in advanced support bundle settings (Support -> Bundles -> Create Support Bundle -> Advanced tab) to ensure that all relevant nodes are included in the selection list.

ORION-194400: Enhanced synchronization between VMS and Uplink to avoid situations where Uplink shows a task as running while VMS reports it as complete.