Backup Job Hangs in 5.0

[turner7577]

Every so often, out of 20 jobs that run each night, i'll have one or two hang after writing just one file and 40192 bytes. it will hang the rest of the night and the only way to cancel it successfully is to right click, choose cancel and reboot the box, it will not cancel out until rebooting the box.

i'm writing to disk, these jobs should fly, especially being diffs, they're tiny!

any ideas?!?!?!

 

[lenski]

Yes this seems to be a common issue with NB on windows. If you look around many of the questions are about this problem. It has nothing to do with the disks. As far as I have seen nobody has supplied an real solution.

 

[GreyWanderer]

First off, I hope you are not running native 5.0




Having been in that boat I can say;

The cancel issue is a known problem and was successfully addressed in MP3.

 

[GreyWanderer]

As far as the original issue with the hang, that could be many things. The client not having enough resources to send the data, network, timeouts, etc. You should be able to get to the root of the hang problem by turning on the bpbkar logging on the client side.

 

[turner7577]

I am running NB 5.0MP3 ... we updated that a few months ago i believe. when doing an update, does just the netbackup server have to be updated or do all the clients also have to be updated?

with the client hangs, its not the same client that hangs each time and it's not every night that one or two hangs. but it happens just so often that it requires a nightly check to ensure a running status of each client.

 

[GreyWanderer]

Sounds more like a network issue, just enough data trickling across so the client timeout does not get invoked. Have you looked at the network side? SAN and TCP if applicable, specifically port utilization and errors?
Have you attempted to kill the PID via cmd line to save from doing a reboot?

Backlevel clients are supported.