To be installed on:  All servers (including clients doing client direct) that  can write to a MSDP storage pool

The following fixes are included in this NetBackup 8.2 MSDP EEB HotFix Bundle.

Etrack 3980925: Enable Auto DR - Cloud DR with manual KMS key transfer

Etrack 3981997: All replications are failing after upgrade to 8.2

Etrack 3980928: Restore of RHV VM fails from the backup image which resides on Amazon with

Etrack 3974756: Backups run slow to MSDP. Intermittently delays between 4-15 minutes starting to perform SQL transaction log backups but when data transfer starts the backup complete successful with about 40 seconds.

Etrack 3983507: Incorrect conversion progress, 99% is shown though there is large a percentage of data containers not converted, or the progress is always 0 before finished

Etrack 3981997: All replications are failing after upgrade to 8.2

Etrack 3993839: EEB of image sharing support for HCP cloud provider
Etrack 3985131: Support stop crc check in fast mode
Etrack 3980502: Don't allow to use same MSDP user from different NetBackup domains
Etrack 3991490: Don't check container existence when commit, esfs will do it
Etrack 3981983: improve ushare performance(for ushare backup performance improvement, need install NBU EEB 3983671)

Etrack 3983210: MSDP spoold deamon fails to start on Access 3340 with 'Too many open files'

Etrack 3979757: [AutoDR]Dirty sst file is not deleted if backup job is canceled
Etrack 3991490: Don't check container existence when commit, esfs will do it
Etrack 3994188: Wide open access shared memory file created by 8.1.2 MSDP media server
Etrack 3995717: NBU restore from u-share backup is 60% drop comparing with regular file restore
Etrack 3985824: Enable bwlimit parameter when do replicate
Etrack 3988744: VLD disable mtstrmd in second STU fragment
Etrack 4000347: Poor deduplication rate during NDMP backup with Dell Unity 300 storage to NBU Appliance 5240
MSDP on NFS
Fix FC hang issue
Make filedel work for both MD5 and SHA2

Etrack 4008599: Not use the MD5 fingerprint in case of unexpected reuse flag change
Etrack 4006469: removed file cause checkpoint failure
Etrack 4012328: Improve the compaction check process to make it easier to use and be more reliable
Etrack 4016575: Improve segment encryption key strength
Etrack 4018989: Limit the error messasge number
Etrack 4019232: Allow only 1 spoold to start at the same time
Etrack 4018886: Old mtstrmd(8.2 or older) can not do cloudlsu backup
Etrack 4019294: Disable pre-download when image is opened with flag  STS_O_DISABLE_PREDOWNLOAD
Etrack 4020699: Don't call ioctl by default
Etrack 4020690: Race condition between load/save header file
Etrack 4020801: Fix a potential vpfsd coredump issue
Etrack 4021113: [vpfsd] vpfsd log has lots of similar ERR: ERROR [33098: [140655629051648]: ]: Failed to Flush all data, error: errno 2
Etrack 4021009: Add lock before loading lmdb for SO list
Etrack 4020164: Add progress info for compaction check
Etrack 4021534: Set time of validity for compaction precheck marker file
Etrack 4021632: Add version for compaction check
Etrack 4021750: handle exceptions for vpfs

Enable nommap for vxfs by default

Etrack 4022222: Fix shareID-DC_group map issue

Enable nommap for more files

Etrack 4023051: Support multi-threading compaction check
Etrack 4022549: Change the option --findgoodbackup to echo more images in catdbutil tool for the msdpcheck tool.
Disable predownload for vmware when bpbrm gets fingerprint align guide through ioctl. In pdplugin, the function pi_ioctl_v9 calls pi_open_image_v10 without flag STS_O_DISABLE_PREDOWNLOAD to disable predownload.
Etrack 4026927: fix memory issue while scanning dc offset from the bin file

Etrack 4029263: Universal Share deduplication Issue in Azure Cloud - MSDP
Fix potential memory leak,if compaction find bin and bhd are mismatch, and bin file has SO/DO meta data, memory leak could happen

Etrack 4033402: Encryption Crawler for JPMC
Etrack 4033123: Ignore uldb error, later will get udcid from spoold
Etrack 4033757: In Glacier spoold may open a uploaded container for writing SO,if container is uploaded to Glacier and is not in esfs local cache, open will fail.
Etrack 4036552: Improve dcscan to dump DOs for the DAT tool
Etrack 4039335: Improve the findkms functionality
Etrack 4038652: For dc miss cases, try to report corrruptions when refdb header corrupt flag was set

Etrack 4040892: 'Dcscan --fixbhd' can cause data corruption in a env with the new SO key generation algorithm

Etrack 4040300: vpfsd KMS encryption support for write path.
Etrack 4031557: vpfs seems dead lock and take checkpoint action hang
Etrack 4032487 : reference for the big file is incorrect
Etrack 4038646 : exception enhancement for adding reference
Etrack 4038770 :need to check the sync references return value during universal share checkpoint to avoid potential data lost issue
Etrack 4037145 :vpfsck for universal share to check consistency between the metadata and data

Etrack 4042607 Universal Shares slow and unresponsive
Etrack 4038864  After running msdpcheck, spad can not startup

Etrack 4048644: Enable compaction check dispite refdb state

enhance fix_bhd, ignore ntotal_rec in bhd header,try read records from bhd file, then update ntotal_rec
Etrack 4058092: handle create file group exception
Etrack 4067149: Don't return and just log an error if shadow file is corrupt
Etrack 4040344: Split the online checking result into small part and send it to spad
Etrack 4042622: Fixed the buffer size overrun caused by large amount of input data
Etrack 4039025: Don't support PDDO and do not try https case, so that nbu web server does not output the url request log which contain plain password
Etrack 4067818: Spad memory usage exceeds 100 GB

Version 35 adds:

Etrack 4080185: Create a new seg for each TAR header
 

For details on the VxFS fixes:

Flex VxFS patch and MSDP EEB bundle
https://www.veritas.com/docs/100049433