Backup Agent crashing on Media Server
Hi All,
We are running BE 2010 R2 with all of the most recent hotfixes through Liveupdate. For the past couple of months (at least) we have been experiencing random job failures on our Exchange off-host backup job that is caused by the remote backup agent on the BE 2010 server crashing. We have 21 jobs configured but the Exchange off-host job is the only one that seems to cause the agent to crash. The off-host job backs up to disk nightly on the media server. I have another job configured to run on the weekends that creates a full backup to disk then to tape that does not experience this problem, but i backup across the network because I needed to ensure 100% stability. I use the off-host to do nightly full's since our server is under constant load 24/7 through the week. You can clearly see it calling it an associated dll with the offhost backup option, but not sure if this is a bug in the BE, a configuration issue, compatibility problem with Equallogic interface driver or what.
I am including the entry from the Event Viewer:
Faulting application beremote.exe, version 13.0.4164.109, time stamp 0x4cffac43, faulting module bedsoffhost.dll, version 13.0.4164.0, time stamp 0x4c2a6ee6, exception code 0xc0000005, fault offset 0x000000000001a0bd, process id 0x17e8, application start time 0x01cc1f1c30a4cb1d
I am at a loss as to where to look. I have deleted and re-created this job in the past, reinstalled backup exec, restarted the server on more than one occassion, uninstalled and reinstalled the remote agent on the exchange server. All the regular troubleshooting steps.
The backup job is about 750gb in size and typically averages around 3500-4000 MB/min. Its pulling from an Equallogic group using multi-pathing and the equallogic vss snapshot software is installed. I have tried changing from automatic to hardware in the advanced disk based backup option for the off-host backup processing with no difference. The server is exchange 2007 SP2 with latest updates and patches upto beginning of April but this problem has existed since before then.
Any assistance would be great!