NBU6 MP3 is out, has anyone started to deploy it yet?

[Alasdair_McQuir]

NBU6 MP3 is out, has anyone started to deploy it yet?

Long list of fixes as well.

[AKopel]

Actually, I'm not sure the it ONLY affect in-line tape copy. It might be the multiple data streams. We do both for our weekly fulls and I haven't tested taking the in-line copy off...
1st problem has to do with these jobs (in-line and/or multiple data streams) not requeuing when they fail for whatever reason. 2nd problem is related to it in that if you manually restart the streams that failed, subsequent jobs (say your incrementals the next day) will not run.

If I were you, to be safe, if you are using any sort of multiple-data stream jobs, I would watch them VERY closely and make sure all expected streams are running every day. Tedious? Yes. but much better than finding out that a D: drive on some important server hasn't run in a week....

[Stumpr2]

Thanks Aaron.

[Alasdair_McQuir]

This sounds very similar to what I am having. I only have multiple datastreams and no inline tape copy. I did try backups with only one stream and they appeared to go ok.

I have an issue with parent jobs starting and then the child jobs do not start. this only happens on file store policies. SQL and exchange appear unaffected.

The only way I have been able to get rid of hung jobs is by restarting the server and then closing the jobs down.

I am also shuting down the AV and seeing how it behaves.

I have been asked todo full logging on vxlogview logging as well as BPBRM and BPTM.

[Anonymous]

Delay in jobs appearing in Activity Monitor

Applied MP3 on master (HPUX IA64) and all media servers, and now any Manual or Immediate Backups seem to delay for exactly 10 minutes before appearing in the Activity Monitor.

Anyone else with this, or seen this before.

[AKopel]

Yes,
This is exactly it.. and what is worse is SOME of the child jobs will start but not all.
It is on only file store policies.

I don't believe shutting down AV will fix it as it is a known issue and we are working with Symantec on it.

[Ed_Wilts]

We just received and installed engineering binaries for nbpem (dated last night) and nbjm (dated June 30). Time will tell if this resolves our scheduling issues.

Cheers,
.../Ed

[Alasdair_McQuir]

Hi Aron, quite right, most of my child jobs do start yet random ones fail.

MP3 has done nothing to fix this as far as I can tell.

All agent policies work fine IE SQL, exchange.

I can't believe we are having so many issues with sucha major product.

[AKopel]

Well,
our 1st issue is NOT resolved yet. Our multiple data stream jobs that had media write errors did NOT requeue. Just went into an incomplete state. I was able to resume them and they will likely finish. I will find out tuesday night if the streams that did this will run or not.

(we have the same engineering binaries for nbpem and nbjm)

Ed,
What company do you work for in St. Paul? I'm at General Mills in Golden Valley.

AK

[Ed_Wilts]

I've got a new engineering binary for nbpem dated Friday at 13:38 but it didn't resolve our problems either. Symantec is still working on it...

It's currently not scheduling jobs that are due, scheduling jobs that aren't due, and not scheduling our destaging jobs at all. Naturally all of it is intermittent (which makes it a lot tougher for the developers working on the problems).

BTW, I'm at Merrill Corporation near the state fair grounds in STP.

.../Ed

[Stumpr2]

uhmmm...how are multiplexed backups restoring. Has anyone experienced media position errors after one of the multiplexed backup jobs fails or is terminated by admin? Have you then tried doing a restore from that tape that experienced the media position error? I think I might have a loss of data situation. 6.0 MP3

perhaps this explains it best
When using Veritas NetBackup (tm) 6.0 through 6.0 MP2, if a multiplexed backup receives an end of media (EOM) message, and before it can get new media, another backup to the same drive fails, there is the potential for data loss.
http://support.veritas.com/docs/282787

but this was suppose to have been fixed with MP3 !!!!Message was edited by:
Bob Stump

[AKopel]

We multiplex almost ALL of our backups and have not had any problems (yet). Although I haven't testing this specific scenario. You are trying to do a restore of a DIFFERENT image on the same tape (different than the one that failed)? Or did the one that failed have checkpoints enabled, finished, and you are now trying to restore THAT image?

[Steve_Froeschke]

We've deployed MP3 last Friday and are watching things closely here. So far, it "appears" scheduled backups are running, but I want to wait a bit longer before we say for sure. We didn't have any problems with the upgrade. Running a Solaris Master server and 5 windows SAN media servers with about 400 clients.

[Alasdair_McQuir]

Support have suggested I turn off NOM which was running on the same server.

I did this yesterday and the server has run fine backups last night.

This could be conincidence so I will be watching it very closely.

Does anyone else have NOM installed to the same server as their master server?

[Stephan_Langbei]

To solve our problems (refer http://forums.veritas.com/discussions/thread.jspa?messageID=4441031쏇 ), we get the MP3-files 5 weeks ago (we have basic support and five critical calls after Install MP2 in march :

vnetd.exe
nbproxy.exe
nbpem.exe
nbjm.exe
nbgenjob.exe
bprd.exe
bpdbm.exe
bpcompatd.exe

and since I have install this files, my backup runs without errors. I install no MP again until I get the ganantie from veritas support, that I don't get errors again like with MP2 - that was no MP, that was an errorpack, an joke from the development.

[Anand_Vidhani]

friday I applied NB 6.0 MP3, I has NB 6.0 MP2. After applying now i am getting these entry in event viewer The "NetBackup Job Manager service terminated unexpectedly. " (nbjm.exe is crashing). Opened the case with veritas. They have given new nbpem.exe , nbjm.exe & process for change !!!

Anyone did this change?. I have master server on windows 2003 SP1

[Ed_Wilts]

> I've got a new engineering binary for nbpem dated
> Friday at 13:38 but it didn't resolve our problems
> either. Symantec is still working on it...
>
> It's currently not scheduling jobs that are due,
> scheduling jobs that aren't due, and not scheduling
> our destaging jobs at all. Naturally all of it is
> intermittent (which makes it a lot tougher for the
> developers working on the problems).

We've received many updates to nbpem since I posted this 10 days ago, and each one has gotten better. The one we put up yesterday seems, so far, to be pretty good. Of course, we've had only 1 night's backups to base this on, but most of the previous nbpem releases would show scheduling issues fairly quickly. Not everything is fixed yet, but it's definitely moving in the right direction.

I would still recommend that anybody at 5.x wait to upgrade to 6.0 (probably until MP4), but if you're already running 6.0, contact support immediately if you're having scheduling issues.

.../Ed

[eric_lilleness]

Sorry, but I gotta ask:

For all you folks having problems:

Windows or UNIX?
Are you positve that you have installed all the patches you need for your installed features?

Thanks, Eric

[Ed_Wilts]

> Sorry, but I gotta ask:
>
> For all you folks having problems:
>
> Windows or UNIX?

Unix. Solaris for me.

> Are you positve that you have installed all the
> patches you need for your installed features?

We're positive - Engineering is in touch with us daily and if we weren't up to date, they would certainly tell us. I don't know how many logs we've sent them - it was 49GB in one day last week.

Like I said, the new releases are getting better, but we're not completely there yet. We still have significant issues with status 90, 200, 80x errors to resolve (along with a bunch of others).

.../Ed

[Anand_Vidhani]

Windows 2k3 sp1 for me..

Today I changed the bin files. Let's see how tonights backups will run..!

Did any one got any new bin files (nbjm & nbpem).

[Alasdair_McQuir]

Yep mine is uptodate, have seen it hang the parent job on the master server its self.

However since I took off NOM from the machine it seems to be running a lot better.