ONBAR woes :-( Advice needed

OK gurus, riddle me this one! 7.31 on solaris 10 nb 6.5.4 Configuring my upgraded Netbackup install, all seemed to be working as expected. However, this morning whilst tweaking the policies for failover i attempted a logical log backup which failed (policy not active) doh! So i activate the policy. I tried again - this time we connect to storage manager and i see the tape being mounted. NB decides that this tape had EMM data on it so is rejected and reqeuest the next available - this turn out to be a freaking cleaning tape !! Somebody in for earache methinks. Anyway good ole NB tells me this tape is full and requests the next - then tries to mount the same bloomin tape again. Soo, enough of this nonsense think i - and kill the job from the db. I add valid tapes to the pool and also to the scratchpool and try again. Whaddaya know - i get this error in the bar_act.log "XBSA Error (BSAInit): That object already exists" Not good - on interrogating sysutils i see there actually IS a job for this log - however this is not reflected in ixbar.<servernum> or the output of onstat -l Now i am pretty much screwed. The job bailed with an "error 13" so how on earth can it be seen as good in the sysutils db???? As of this moment i am up a certain creek without a padddle - anybody seen this kind of thing before? I don't really want to start hacking sysutils and breaking RI etc etc but am pretty stumped as to what to do.

BRYAN PURKIS wrote: snip.... > > I add valid tapes to the pool and also to the scratchpool and try again. > > Whaddaya know - i get this error in the bar_act.log "XBSA Error (BSAInit): > That object already exists" > > Not good - on interrogating sysutils i see there actually IS a job for this > log - however this is not reflected in ixbar.<servernum> or the output of > onstat -l> > Now i am pretty much screwed. The job bailed with an "error 13" so how on > earth can it be seen as good in the sysutils db???? > > As of this moment i am up a certain creek without a padddle - anybody seen > this kind of thing before? I don't really want to start hacking sysutils and > breaking RI etc etc but am pretty stumped as to what to do. > I do think the entry in sysutils is NOT the problem. The entry probably does reflect the failed attempt to archive the log. The XBSA Error comes from your storage manager. I think that the object (i.e. failed attempt to archive the log) made it to the index of your storage manager somehow (although it does not exist) and now it complains that it already exists. Maybe the entry has to be removed from the SM index manually. HTH Tilman

Hello. We had to debug some netbackup processes also, and what helped was the following actions: I adjusted the BAR_DEBUG to 9 in onconfig file, I read the official documentation (a pdf file from symantec about NetBackup for Informix), and activated also the recording of netbackup client logs on linux. Since then, I was able to trace the errors (mainly on our BAR_DEBUG_LOG file, until it raised something strange). Hope this helps on your situation. Regards. Alexandre Marini Analista de Tecnologia da Informação - DBA SEFAZ-MS / SGI-UIMP / Sistemas IBM-Informix tilleul17@web.de escreveu: > BRYAN PURKIS wrote: > snip.... > >> I add valid tapes to the pool and also to the scratchpool and try again. >> >> Whaddaya know - i get this error in the bar_act.log "XBSA Error (BSAInit): >> That object already exists" >> >> Not good - on interrogating sysutils i see there actually IS a job for this >> log - however this is not reflected in ixbar.<servernum> or the output of >> onstat -l>> >> Now i am pretty much screwed. The job bailed with an "error 13" so how on >> earth can it be seen as good in the sysutils db???? >> >> As of this moment i am up a certain creek without a padddle - anybody seen >> this kind of thing before? I don't really want to start hacking sysutils and >> breaking RI etc etc but am pretty stumped as to what to do. >> >> > > I do think the entry in sysutils is NOT the problem. The entry probably > does reflect the failed attempt to archive the log. > The XBSA Error comes from your storage manager. I think that the object > (i.e. failed attempt to archive the log) made it to the index of your > storage manager somehow (although it does not exist) and now it > complains that it already exists. > Maybe the entry has to be removed from the SM index manually. > > HTH > Tilman > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > > >

Bryan, You may already be past this problem, but you might want to check out the onsymsync utility to sync your sysutils to your netbackup catalog.... it might help. Here's a link to v10 doc, .... oh, just saw you are on 7.31, not sure onsymsync worked then... when I was on earlier versions, the onsymsync utility didn't' really seem to work so we looked at the scripts it runs and reverse engineered the scripts to work for us... http://publib.boulder.ibm.com/infocenter/idshelp/v10/index.jsp?topic=/co m.ibm.bar.doc/barmst193.htm -----Original Message----- From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of BRYAN PURKIS Sent: Friday, July 17, 2009 6:38 AM To: ids@iiug.org Subject: ONBAR woes :-( Advice needed [16430] OK gurus, riddle me this one! 7.31 on solaris 10 nb 6.5.4 Configuring my upgraded Netbackup install, all seemed to be working as expected. However, this morning whilst tweaking the policies for failover i attempted a logical log backup which failed (policy not active) doh! So i activate the policy. I tried again - this time we connect to storage manager and i see the tape being mounted. NB decides that this tape had EMM data on it so is rejected and reqeuest the next available - this turn out to be a freaking cleaning tape !! Somebody in for earache methinks. Anyway good ole NB tells me this tape is full and requests the next - then tries to mount the same bloomin tape again. Soo, enough of this nonsense think i - and kill the job from the db. I add valid tapes to the pool and also to the scratchpool and try again. Whaddaya know - i get this error in the bar_act.log "XBSA Error (BSAInit): That object already exists" Not good - on interrogating sysutils i see there actually IS a job for this log - however this is not reflected in ixbar.<servernum> or the output of onstat -l Now i am pretty much screwed. The job bailed with an "error 13" so how on earth can it be seen as good in the sysutils db???? As of this moment i am up a certain creek without a padddle - anybody seen this kind of thing before? I don't really want to start hacking sysutils and breaking RI etc etc but am pretty stumped as to what to do. ************************************************************************ ******* Forum Note: Use "Reply" to post a response in the discussion forum.