Veritas Bug ID: ET 4140863

Version:  NetBackup 10.3.0.1

Installation Location: Server (MSDP Cloud media server)

Problem Description: MSDP Direct Cloud Tiering EEB bundle with the latest fixes for 10.3.0.1

Fixes included:

ET 4149275: Fix core dump in Azure env caused by expired token.
ET 4136545: Fail to set WORM retention in backup in AWS_RVLT.
ET 4149972: Fail to Edit lock duration of worm disk volume, Reconfiguration failed for storage server.
ET 4156890: Cannot list and create bucket for Ceph s3 in WebUI volume configure.
ET 4156337: Avoid crash in ocsd if blob.versionID is nil.
ET 4135259: Retry failed delete requests for cloud storage. Delete requests may fail for a short time if connectivity to the cloud is interrupted or for a number of other reasons. Stop retrying after a reasonable number of attempts to delete. Remove container from our databases if we are unable to delete it after a reasonable amount of time. Do not remove the record in ccdb for CloudCatalyst migrated objects if the delete fails so that we are able to retry.
ET 4140926: Do not track delete requests for retry in read_only mode. Avoids hang in image sharing server where dcdb does not exist.
ET 4156139: 'catdbutil --worm disable' command should handle AWS STS token expired error and repeatable.

ET 4162377: Oracle Backups WORM Images are failing with error Code 252.
ET 4161529: Duplication from MSDP to MSDP-C is failing with an error message OpenStorage WORM lock error (88).
ET 4158995: Failed to disable retention lock for image.

ET 4168643: Failed to configure WORM volume for Azure China.

ET 4131873: Fail to create bucket with default retention in GCP.
ET 4179058: 'msdpcldutil update live' failed with 'Access Denied' error for AWS S3 bucket on Recovery Vault.
ET 4180195: Failed to switch mode to Compliance from WebUI with ERROR 'Failed to update the retention period of the object '<bucket>/<disk volume name>/lockdown-mode.conf'
ET 4181448: Retry Azure server errors on upload with exponential delay to address a specific problem with very large duplication jobs.
 

Version 14 Adds:

ET 4189822: Assume upload success if blob is in archive tier. On retry after connection reset by peer, if Azure returns the status that the blob already exists in  the archive tier then assume it was uploaded successfully.
ET 4182945: Renew the RVLT SAS token for any status forbidden error for RVLT Azure. It appears that the 'specified time frame' phrase is not always in the error message from Azure.
ET 4180970: Fix for AWS Restore from Archive, this would be a special case of RVLT where the restore and status request taking too long to execute due to various  networking/workload reasons. Fix out of index error in handleWarmCheck, if the objects are less than batch size which is 500 objects.
ET 4192354: Spawning multiple go routines to expedite the restore/stat requests.
ET 4182945: Fix a bug during azure warming.
ET 4190903: List non-worm buckets fails with 'Only IAM users are supported with object lock enabled buckets'.
ET 4192354: Initiate restore and then add sleep during restore, update warmed variable.
ET 4194941: Avoid crash in ocsd, when VersionId is nil.
ET 4192354: Add wait group on the warm go routines and retry warm if any one of the batchFileWarm routines fails due to expired SAS.
ET 4192354: Separated restore and stat operations so that we can efficiently retry for warm failure and fixed the OCSD crash when reading from multiple senders.
ET 4195289: Enforce the use of custom endpoints for WORM volume creation.
ET 4195652: Resolve performance issues in cleanup on Hitachi local/private cloud. Skip WantUntilObjectNotExists for hitachi-csl.
ET 4204404: HeadOject gives objectLock properties, check it before trying to set retention on container with retention. Avoid an excessive number of 403 status codes from API calls when setting shorter retention on existing locked WORM container with longer retention already set. Occurs in environments with policies having different retention levels.