Hi Iman,

What build of 8.2 SP3 did you upgrade? I would recommend go with the latest cumulative 8.2.3.2 build 1121. It may help you get past this error.

Regards,

Akshay Bhumkar

Currently using 8.2.3.1. Would like to look for some fix before upgrading to 8.2.3.2.. if any.

Java is not involved with the process just with the visual presentation of the result.

Of course you can go though all the logs but I think you better start with more verbosity.

May I suggest you run the group from the command line using this command:

     savegrp -v -G

This should tell you more.

This may depend where you're getting your information.  Specifically, what are you looking at to know that each client backed up or that the savegroup was successful? 

There was a problem historically with the TCP Chimney implementation in Windows (*maybe* not Windows specific, but I know for sure it existed there) where saves would complete to media successfully, but during the shutdown of the TCP connections between client and server some packets would not be acknowledged.  This resulted in savesets which were complete, successful backups according to DPA (which checks saveset status), but failed backups according to NMC due to a communications error.  Depending on your point of reference, all are correct.

There's also a potential discrepancy if you're looking at DPA reports which will show failed savesets which are rerun successfully.  Again, it's right that there was a failed saveset, but doesn't easily show that the saveset was also successful.

Hello,

First of all there should be any issue with java incompatibility.

But to troubleshoot the you might have to investigate more and need to find out belwo concern -

• i) from where you are actually checking the failure log , is it form dameon.log or from some other reporting tool like DPA
• ii) Do you have any additional message on /nsr/logs/sg/
• iii) Is your savegrp taking care of index backup also, if yes then ensure no issue with index backup.
• iv) Is this for a specific savegroup or all savegroup? if for specific try to find out the savegrp option, and also notification setting.
• v) Always recommended to run a manual savegrp with debug mode and see the output where exactly its stacking -

     savegrp -vvv -D9 savegroup_name > /tmp/savgrp.log

Thanks guys for the reply. The issue is now fixed.

The problem was with gss authentication for some of the clients in the group. When the server rejected the credential it causing the group failed. Fixed the issue by clearing peer information on both server and clients side.