Veritas Bug ID: ET 4122719

Problem Description: NetBackup 10.2.0.1 MSDP Preferred EEB Bundle

Version:  NetBackup 10.2.0.1

Installation Location:  Media Server, Client

If client side deduplication is used, this EEB should be installed on the client and MSDP server. All servers that are MSDP storage pools or can write to a MSDP storage pool should also have this EEB installed.

This Hotfix contains multiple Emergency Engineering Binaries which fix problems with MSDP Servers, Flex, and Flex WORM Appliances.

Fixes Included in this NetBackup 10.2.0.1 MSDP EEB Bundle:

ET 4124834: Backport kek_tag_reporting and crcontrol createnewkek option, includes Incident 4118545, Incident 4116618, Incident 4113795, Incident 4115764.
ET 4120242: cleanup ushare backup snapshot copies may cause memory leak
ET 4118121: add try catch for file system release
ET 4118480: usharecli read all OST image from C1F1, C1F2, C1F3 ... C1FX, not just read the first C1F1 image
ET 4118349: backup and background task deadlock
ET 4119452: After ushare backup, loop device might be hold by mtstrmd and other processes and could not release
ET 4125603: Add engine down check to improve scache hookup performance
ET 4127588: Check if the containers are in recover mode nodes.
ET 4127761: snapshot may mount incorrect head volume 
ET 4130738: file path may be missing while rename a file in UShare
ET 4130516: set vpfs dsid encryption when dsid 0 is set during initialization.
ET 4130860: Fix oracle copilot with ushare backup from snapshot fail with status 40/6 issue (errno 9 when dump file's SO list)
ET 4127577: fix out of memory issue.
ET 4130976: Close the DC iterator when all entries are accessed.
ET 4118234: Close dc data file to clean up the open files and free memory
ET 4122249: The performance issue of vpfsd instances from different container instances are bound to the same set of CPU's on Flex
ET 4128953: spoold's loopback noblocking socket stop work in windows 2019
ET 4132111: Too many logs in online check log files
ET 4128845: Spad hang due to lock leak 
ET 4131769: devname in /etc/fstab, use .vxblkdevuuidmap to get uuid
ET 4128638: Disable SMBv1 due to 'Deprecated Protocol SMBv1 Supported'
ET 4123495: High memory RSS usage of vpfsd process after running Ushare dump and backups for several hours, cause pod/spoold restart
ET 4126907: compaction needn't call loadIndexFromContainer() if s-p cache is enabled
ET 4138610: Media Server instance off-line due to /mnt/nbdata run out of inodes
ET 4104218: ET4104218 - Media Write Error(84)
ET 4139144: Ushare backup reports no such file or directory during ushare dump and backup concurrently.

ET 4139944: porting:check errno instead the return value and if path noexist try to creating it    
ET 4141039: NetBackup Flex WORM Universal Share: incomplete MSDP cleanup or poor deduplication - PO size incorrect

ET 4133206: Ushare accelerator crashes while doing multiple types of operations against it
ET 4136443: Porting fix to correct the create folder check
ET 4124639: During scalability case running, found ushare access hang issue
ET 4139748: chdir() should not be called from DeleteTree() especially in a multi-threaded process, otherwise CWD of the process would be wrong forever.
ET 4142962: vpfsd hang during concurrent dumping data to ushare from 2 clients, the following ushare backup also hang

ET 4112468: Fix request getLsuBucketRetention is sent to spad too often issue.
ET 4143329: Adjust the online check batch sizeIncident:

ET 4134616: vpfs popen to close parent file handles before executing child process
ET 4133598: Ushare backup fails on 40 error, hangs at mkfs
ET 4147534: move extent failure due to not enough space
ET 4144312: Fix coredump of dcscan

ET 4153909: bad recovering file causes vpfsd cannot be repaired autmaticly
ET 4120485: Disable image data container retention lock.
ET 4153152: ignore filename having surrogates
ET 4113961: Add option in pd.conf to allow MSSQL backup using multi-stream
ET 4152661: Use token credential to disable WORM image in Recovery Vault cloud provider.
ET 4132244: Fix delete share may fail if some shares are initializing
ET 4146413: update im if tir or data vector is empty
ET 4145149: Improve SQL backup performance in NBFS
ET 4110247: Validate host only if secure comms is enabled
ET 4153102: fix boost interprocess security issue. use programdata folder
ET 4118717: support independent disk for instant access vm
ET 4156139: 'catdbutil --worm disable' command should handle AWS STS token expired error and repeatable.
ET 4156951: Credentials print error.
ET 4156775: increase inode space usage from 25% (default) to 80%
ET 4145688: limit malware scan retention to 3 days
ET 4127387: UShare dump failed and VPFS reported lots of stack trace ERROR, after rename  failure
ET 4153102: fix boost interprocess security issue. use programdata folder

ET 4158688: Improve cata2map logs
ET 4158977: Fix the vpfs_mounts hang issue when /etc/resolv.conf is empty
ET 4159164: Missing /var/log/nginx on WORM instance
ET 4159647: Flush race condition observed while dumping data to UShare accelerator for RDS PGSQL Database
ET 4156433: After upgrade to Access 8.2 NBU jobs start but do not progress
ET 4160369: Universal share backup failing with status 40, vpfs logs show 'size should not be 0, but its size is 0'
ET 4162644: UShare backup hang 
ET 4162456: Vpfs stability enhancements
ET 4161170: Unable to reconnect to MSDP Storage after upgrade to NBU 10.4Unable to reconnect to MSDP Storage after upgrade to NBU 10.4
ET 4159340: Split the init into two parts, create threads only if all other components have been initialized
ET 4163635: Set default WORM min = 1 day, max = 1 year in spadb
ET 4162340: Notify memory allocator to release memory if it reserve too much memoryCRQP and crcontrol can ask tcmalloc and glibc to free reserved memoryIncident:
ET 4164500: Call prlimit to let spoold be able to mlock the whole FP cache
ET 4158136: Log more details for I/O error
ET 4164426: After parallel duplication to MSPD-C WORM only first copy created gets as WORM locked

Version 19 includes/adds:

ET 4181061: 'crcontrol --memallocator release' hung
ET 4178785: Poor backup performance after migration to 5360 Flex HA appliances
ET 4177901: Fix the invalidation of config param OPTDUP_BANDWIDTH when param BandwidthLimit is zero.
ET 4175374: fix chown for folder: if user or group is not provided, just keep original value.
ET 4173259: let merge continue if it has been canceled too many times