Enhancements in 5.2.2

Added an ability to manage a subset of VAST OS configurations in a new package named vast-utils, which is upgraded during NDUs. The vast-utils package includes cluster networking configurations and host-level services.

Starting with VAST Cluster 5.2.2, most VAST OS configuration updates will be delivered simply by upgrading the vast-utils package, effectively eliminating the need to perform a full VAST OS upgrade and a node reboot.

The currently installed version of vast-utils is displayed in the VAST-UTILS Version column for each node in the Infrastructure pages in VAST Web UI, as well as in the VAST Utils Version column of cnode list , cnode show, dnode list, dnode show command output in VAST CLI.

ORION-260525: Added an ability to specify the domain name when entering EKM servers in VAST Cluster Install.

ORION-220688: For clusters with InfiniBand internal networking, enhanced OS upgrade routines so that the OpenSM master node gets upgraded prior to the VMS node.

ORION-241294: For clusters with InfiniBand internal networking, updated the way VMS maintains OpenSM configuration parameters so that on the first start of the OpenSM service, the parameters are taken from the OpenSM configuration file (/etc/opensm/opensm.conf).

In addition, the opensm.conf will include two more parameters, max_wire_smps 96 and max_wire_smps2 96, which set the maximum number of SMPs that can be sent in parallel.

ORION-237091: Increased the maximum allowed number of tenants per cluster from 2048 to 4096.

ORION-207626: Users that are associated with a VAST tenant are now able to view virtual IP pools that are not associated with a particular tenant(s).

ORION-256797: Added metrics that help analyze QoS credits left per view. To display the metrics in VAST Analytics, create a new customized analytics report and include the metrics starting with the word 'Burst' for the View object.

ORION-253117: Made updates to automatically set a default value for credits when a maximum static limit is set.

If no credits are explicitly defined, setting a maximum static limit will cause the corresponding credits to accept a default value. The credit default value will be the maximum limit multiplied by 4. For example, if you set the maximum allowed read bandwidth to 500 and do not specify any value for the read bandwidth credit, the read bandwidth credit will automatically be set to 2000.

ORION-250317: The following limitation no longer apply:

In case of a CNode HA event, the cluster needs a few minutes to retrieve connections from the failed CNode. During this time, the user will be served at the user's S3 connection limit less the number of connections that were provided by the failed CNode.

ORION-240658: The enable-nfs-return-open-permissions setting of a view policy will now be honored with all security flavors. Prior to this change, the setting was ignored if the flavor was S3 Native.

When this setting is specified, the NFS server returns 777 permissions (access is allowed for everyone) for all files and directories when responding to client-side access checks. Otherwise, the server would return 700 permissions (access is allowed for the bucket owner only). The 777 permissions are needed for Windows NFS clients to be able to access S3 buckets where they are not the bucket owner.

For new installations, the enable-nfs-return-open-permissions setting is by default enabled for all security flavors. For upgraded clusters, the setting is honored with NFS security flavor only. If you want to alter the behavior for your cluster, contact VAST Support.

ORION-246616: Enhanced SMB permission handling to avoid dropping SMB generic access rights (GENERIC_ALL, GENERIC_WRIRE, GENERIC_READ, GENERIC_EXECUTE) when converting SMB permissions to VAST permissions (which the cluster uses internally to perform access checks).

Prior to this change, these ACEs could be dropped when performing operations such as running a bulk permission update on a view controlled with SMB or Mixed Last Wins security flavor.

The new logic is by default disabled. To enable the enhancement on your cluster, contact VAST Support.

ORION-187569: Added support for S3 trailing checksums.

VAST Cluster now recognizes that an S3 request includes a checksum and handles the checksum separately from the uploaded content. Note that VAST Cluster does not perform checksum verification.

ORION-231931: Added support for the If-None-Match header in GET and HEAD requests as part of S3 conditional reads. The VAST cluster returns the object only if its ETag differs from the client-specified value; otherwise, it responds with a NotModified status.

Note

When the If-None-Match header contains a list of ETags and this list includes a matching ETag, the VAST cluster returns status code 200 (OK).

ORION-241238: Updated protected path validations to allow for creation of protected paths on a subdirectory of an S3 Bucket view for local protection policies.

ORION-210370: When creating a local user, you are no longer required to supply the NFS uidNumber attribute value for the user. The UID field in VAST Web UI (User Management -> Users -> choose to create or modify a user) can now be left empty; the --uid parameter on the user add command is now optional. Thus you can create users intended only for S3 access that do not need an NFS uidNumber to operate.

ORION-263129: VAST Prometheus Exporter provides new metrics that help detect latency in client socket reads.

ORION-258426: VAST Prometheus Exporter lets you export a metric that shows the amount of active NFS-over-RDMA connections.

ORION-226846: The variables that you can insert into the webhook JSON when configuring a webhook (in VAST Web UI: Settings -> VMS -> Webhooks) include two new timestamps:

ORION-216345: VAST Prometheus Exporter lets you export metrics to analyze data reduction by reduction type (compression, deduplication, similarity).

ORION-213509: Added various metrics to VAST Prometheus Exporter that help analyze cluster's metadata usage.

ORION-250363: Starting with VAST Cluster 5.2.2, requests to the VAST REST API /tenants/ endpoint that contain the vippool_ids parameter are ignored. To associate a virtual IP pool with a tenant, pass the tenant_id parameter to the /vippools endpoint.

ORION-238151: Added support for 3.2TB SCMs and 122TB SSDs.

ORION-247025: Updated high ambient temperature event definitions to lower the thresholds that trigger the alarm.

Rack-level resiliency enhancements: