Networker 7.6 - 71909:save: Failed to close save session

Hi,

This error message doesn't really tell us much.  Your best bet is run the save command on the client itself and get the output of that.  For example:

save -s [server_name] -vvvvv -D9 /etc/hosts > testsave_out 2>&1

This should give you more information around where this is failing exactly.

You can do a backup of something small to test if that works and, if it does, move onto the actual data that you are trying to backup.  It's important to first establish whether the problem is related to the communication between the machines or not, and if it is, whether it's the initial communication that fails (probe failure or even small backup fails) or whether it only fails on certain data or after a certain period of time.

-Bobby

Thanks Bobby,

I tried to do a backup for a small folder on a client but still failed.  But luckily I found out the cause finally.

The reason is that the target media is a D2D device. Though it's not full but the entries on NAS share disk is reaching 25000.  Too many single save sets set on clients.

We've been receiving this same error for some time on various clients.  We rerun the group and this save set typically completes successfully.  It's always an index: save set which writes to AFTD.

Can you elaborate on the issue regarding "the entries on NAS share disk is reaching 25000.  Too many single save sets set on clients?"

I, too, am seeing that error.  I ran a save with the following command:

save -s labhydra -vvvvv /swlib > testsave.log 2>&1

and here are the final lines in the very large file:

6679:save: Connecting directories...
uasm -s /
4596:save: directory contents for /:
`media/'        fileid=29229057
`test/' fileid=52232193
`home/' fileid=26705921
`..'    fileid=    2
`swlib//'       fileid=    0
`labhydra-dd2/' fileid=27099137
`var/'  fileid=40894465
`tftpboot/'     fileid=33521665
`misc/' fileid=    0
`jre/'  fileid=42762241
`lib/'  fileid=31293441
`net/'  fileid=    0
`dev/'  fileid=    0
`lib64/'        fileid=40665089
`srv/'  fileid=38993921
`bin/'  fileid=31752193
`tmp/'  fileid=22183937
`.autorelabel'  fileid=98305
`sbin/' fileid=9207809
`proc/' fileid=    0
`sys/'  fileid=    0
`mnt/'  fileid=6979585
`selinux/'      fileid=20578305
`usr/'  fileid=4423681
`root/' fileid=5439489
`.autofsck'     fileid=98306
`opt/'  fileid=39190529
`boot/' fileid=    0
`lost+found/'   fileid=   11
`etc/'  fileid=50593793
`nsr/'  fileid=37945345
/: fid =
71909:save: Failed to close save session: Save session with saveset id da0fa65e-00000006-2bea468c-4eea468c-01d61b00-70091021 terminated

I believe I started seeing that after I upgraded to NetWorker for Linux V7.6.3.

Thoughts?

Thanks

tl

Hi,

How long did the backup take?  Is the failure consistent (or consistent for all backups that take a certain period of time)?  What are you backing up to?  There have been some issues reported around this area relating to timeouts at various levels.

-Bobby

My errors are on Windows 2008 R2 server and clients.  In recent history, it has only occurred when backing up the index for file servers.  All indexes write to an AFTD on an Data Domain. The backups took 2:02 hrs.(FS1), 7:34 hrs.(FS2), 6:37 hrs.(FS3), and 5:52 hrs.(FS2).

Hi,

You might want to check out on the DataDomain side what the inactivity timeout is and make sure you have the latest DDOS version.  Apparantly there were some timeout issues with some older DDOS versions that cause this sort of issue.  Also, if you're using DDBoost, you can tune the OST_ABANDON_TIMEOUT.  You'll find details in this support article.  I know the environment described in the article does not correspond to yours, but the root cause of the problem is likely to be the same.

-Bobby