Showing results for 
Search instead for 
Did you mean: 

Backup Exec 11 and Job Engine

Level 3
Ugh!  What have you guys done to the job engine to make it so fragile.  We've been using BE for years and just upgraded to version 11.  Mind, I said upgrade and version 10d was working perfectly with all these same jobs.
1)  BE 11 will not do offhost backups!  It crashes the job engine at the start of the job.  The job I have backups up our exchange 2003 mailbox instances (clustered) using Vertas Storage Solution 4.3 MP2.  The job has a prejob script that snapbacks the volumes so that BE can do it's own snapshot and splitting/export of the disk group.  It's not the prejob script because I've even removed it and tested having a free disk mirror set ready for the BE snapshot.  When the job starts, it say preprocessing and nothing happens.  Looking in the veritas storage solutions enterprise admin, I don't see anything occurring suggesting that BE hasn't issued any snapshot request.  After a minuted the job engine service crashes and restarted and the job changes to recover and queued.  I have end up restarting the BE services because the drives in the LTO array are listed as being in use from the previous crashed job.  This is not due to anything that I'm doing different as we've been using BE10 and Veritas storage solutions to do offhost backups for more 18 months now so I'm pretty familiar with what is going on and how it all works.  This is a problem with BE 11 and see so many others complaining about Job engine crashes and exchange it must be BE11.  I've tried the work arounds mentioned in others complaints (disabling GRT and consistancy checks) and they dont resolve the issue but then mine involves the offhost condition.  If I do a regular exchange backup, then yeh I can backup our exchange environment at a cost of several hours.  We have a 6 node cluster with 5 exchange instances average about 1000 mailboxes each so you can imagine the time involved here.
2) With BE11 the backup selection list seems to have a maximum number of objects that can be added.  I had a list that included all the system state backups of all our production servers.  When I open it, it's blank.  If I do a summary, it shows everything.  If I try to create a copy, I get an error with the title of "error creating select" and message of "the parameter is incorrect.".  If I try adding a system to the list, the list only has that new selection and everything is lost.  So I tried recreating the list from scratch.  The first one took 32 hosts that consisted mostly of c, system state and shadow copy components.  The secont one took 37 of the same.  If I try to add anything more to them, and click ok, the dialog flashes for a second by remains.  it only closes if I cancel or remove selections.  If I take those two list and try to merge into one new list, both lists import into the new list but it won't close unless I cancel or remove hosts.
Please tell me that there is a fix for the job engine issues.  The reason we upgraded to BE11 was because we're going to be tasked to backup Oracle on Linux and want to use the new oracle agent for online backups of Oracle.  I'm sure that if I downgraded back to v10 on our primary BE host I'd resolve this issue but I doubt it could be used to manage v11 BE media servers.
I'm posting this but I will more than likely be calling support directly in the next couple days if I don't see a good response.

Level 3
Basically, don't upgrade to verison 11. We had a lot of trouble and now we just uninstall any previous version and do a regular install of Backup Exec 11. This is after one attempt at upgrading and spending 4 hours on the phone with Symantec and having them finally say to uninstall the old version.
The selection list issue is a "feature." It's limited to 154 elements. The fix is here:

Level 3
Thanks for the note on the selection list.
So what you saying is that this issue is due to upgrading and not a "clean" install?  If I was to do a clean install on both my primary and media server how do I export and import my jobs and policies because that is a lot of work lost.   I saw posts about corrupt catalogs and tried to apply suggestions in those posts but it didn't work.  I began testing the Oracle agent for windows and linux and when I submit a job to backup an oracle instance on linux the job engine crashes on that also as soon as it changes to preprocessing.