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]

We will be deploying it today or tomorrow.... We are working with support on some other issues and will already be deploying some custom binaries on top of MP3...

[Alasdair_McQuir]

Have just deployed this, will let you know how it goes.

[Alasdair_McQuir]

What issues have you got above MP3?

[Troy_Schuler_2]

We are downloading it now and will patch today.

Our issues are:

Vault running on excluded day (i.e. 4th of July); Vault job failing with error code 288 or 287 because it won't spit the Vault catalog out of the tape drive and into the Media Access port.

Exchange backups fail now with a 228 complaining about the image catalog. <--scaryMessage was edited by:
Troy Schuler

[AKopel]

One bug (which is not fixed in MP3) is when we have multiple data streams with in-line tape copy. If one of the streams (say a D: drive) fails for some reason (media write error etc) That job does not retry. It just goes into an incomplete state and never tries again. To make matters worse, even if you manually resume or retry the stream (in which case it will finish fine) The next day's backups will run but not for that stream. So it is very hard to troubleshoot.

As an example:
We have a client that has C:, D:, E:, and F: drives.
Multiple data streams with inline tape copy on the weekends.
During the weekend the E: drive might fail because of a media write error or otherwise.
We catch it and resume the job (because it won't retry on it's own)
Job finishes and we 'think' we are ok.
On Monday night, jobs kick off as usual, and everything looks ok, but if you look close, only C: D:, and F: streams were run. E: is nowhere to be found.

So we have 1 of two options:
1) look every morning and verify that all streams have run (not easy)
2) delete the pempersist file every day (better option)

Like I said, we are getting a binary to put on top of MP3 today that is supposed to fix these 2 issues.

AK

[AKopel]

One bug (which is not fixed in MP3) is when we have multiple data streams with in-line tape copy. If one of the streams (say a D: drive) fails for some reason (media write error etc) That job does not retry. It just goes into an incomplete state and never tries again. To make matters worse, even if you manually resume or retry the stream (in which case it will finish fine) The next day's backups will run but not for that stream. So it is very hard to troubleshoot.

As an example:
We have a client that has C:, D:, E:, and F: drives.
Multiple data streams with inline tape copy on the weekends.
During the weekend the E: drive might fail because of a media write error or otherwise.
We catch it and resume the job (because it won't retry on it's own)
Job finishes and we 'think' we are ok.
On Monday night, jobs kick off as usual, and everything looks ok, but if you look close, only C: D:, and F: streams were run. E: is nowhere to be found.

So we have 1 of two options:
1) look every morning and verify that all streams have run (not easy)
2) delete the pempersist file every day (better option)

Like I said, we are getting a binary to put on top of MP3 today that is supposed to fix these 2 issues.

AK

[Alexandre_Rappo]

Check out the TechAlerts - Not just SCARY...
Subject: Symantec Technical Advisory
Dear Valued Symantec Customer,

This is to inform you that Symantec Enterprise Technical Support has just issued a Software Alert. This is a critical technical issue for:
-Veritas NetBackup (tm) Enterprise Server 6.0
-Veritas NetBackup (tm) Server 6.0


Symantec is issuing this Software Alert to inform you that Veritas NetBackup (tm) Enterprise Server/Server 6.0 Maintenance Pack 3 (MP3) was just released June 30th, 2006. Because product quality and responsiveness to customers are consistent Symantec Corporation hallmarks, we are alerting you to some rare but potential data loss issues that are formally resolved with NetBackup 6.0 MP3. These issues were only recently discovered and are as follows:

1. If retention mappings are enabled via the GUI interface in Veritas NetBackup (tm) for Vault 6.0, 6.0 MP1, and 6.0 MP2, and the default retention levels have been modified, vaulted images may expire at a different time than expected and may lead to possible data loss.
http://support.veritas.com/docs/282996

2. When using Veritas NetBackup (tm) 6.0, 6.0 MP1 and 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

3. All Veritas NetBackup (tm) 6.0 logging should reside on a dedicated file system to prevent corruption of the NetBackup Enterprise Media Manager database when a disk full condition occurs.
http://support.veritas.com/docs/282805

4. An issue identified with the resuming of failed jobs using the Checkpoint Restart option can lead to a loss of data due to missed backups when using Veritas NetBackup (tm) 6.0 Maintenance Pack 2 (MP2).
http://support.veritas.com/docs/283180

What you need to know as a NetBackup Administrator is that these issues mentioned herein are resolved with NetBackup 6.0 MP3. Symantec strongly recommends the best practice of always being at the latest patch. If you cannot implement 6.0 MP3 immediately, Symantec highly recommends implementing the workarounds provided in each TechNote if you are running into any of the issues described above.

If you have any questions or concerns about any of these issues, please contact Symantec Enterprise Technical Support.

Thank you for choosing Symantec.


Sincerely,
Symantec
_________________________________________________________

[Ed_Wilts]

We installed it this morning since 6.0MP2 was such a disaster. So far, it's looking better than MP2 but we're not convinced it's fixed everything yet. We'll know more in a day or two.

We had significant scheduling issues with MP2, even with engineering patches. Many jobs simply weren't being scheduled - no failures, just not being scheduled.

.../Ed

[Alexandre_Rappo]

Check out the TechAlerts - Not just SCARY...
Subject: Symantec Technical Advisory
Dear Valued Symantec Customer,

This is to inform you that Symantec Enterprise Technical Support has just issued a Software Alert. This is a critical technical issue for:
-Veritas NetBackup (tm) Enterprise Server 6.0
-Veritas NetBackup (tm) Server 6.0


Symantec is issuing this Software Alert to inform you that Veritas NetBackup (tm) Enterprise Server/Server 6.0 Maintenance Pack 3 (MP3) was just released June 30th, 2006. Because product quality and responsiveness to customers are consistent Symantec Corporation hallmarks, we are alerting you to some rare but potential data loss issues that are formally resolved with NetBackup 6.0 MP3. These issues were only recently discovered and are as follows:

1. If retention mappings are enabled via the GUI interface in Veritas NetBackup (tm) for Vault 6.0, 6.0 MP1, and 6.0 MP2, and the default retention levels have been modified, vaulted images may expire at a different time than expected and may lead to possible data loss.
http://support.veritas.com/docs/282996

2. When using Veritas NetBackup (tm) 6.0, 6.0 MP1 and 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

3. All Veritas NetBackup (tm) 6.0 logging should reside on a dedicated file system to prevent corruption of the NetBackup Enterprise Media Manager database when a disk full condition occurs.
http://support.veritas.com/docs/282805

4. An issue identified with the resuming of failed jobs using the Checkpoint Restart option can lead to a loss of data due to missed backups when using Veritas NetBackup (tm) 6.0 Maintenance Pack 2 (MP2).
http://support.veritas.com/docs/283180

What you need to know as a NetBackup Administrator is that these issues mentioned herein are resolved with NetBackup 6.0 MP3. Symantec strongly recommends the best practice of always being at the latest patch. If you cannot implement 6.0 MP3 immediately, Symantec highly recommends implementing the workarounds provided in each TechNote if you are running into any of the issues described above.

If you have any questions or concerns about any of these issues, please contact Symantec Enterprise Technical Support.

Thank you for choosing Symantec.


Sincerely,
Symantec
_________________________________________________________

[Alasdair_McQuir]

The patch was applied yesterday, it has made no difference to the backups what so ever.

They are the parent job still starts and the child jobs do not kick off.

Am escalating this with Veritas most urgently.

[Alasdair_McQuir]

I now have even more strange behaviour.

I am unsing mutiplexing and multistreaming.

I have a number of parent jobs that started at midnight and now at 10.00am the child jobs have started but for a weekly shcedule ratehr than a daily schedule.

Its like somthing out of the fortean times

[Stumpr2]

I've jumped on to 6.0 MP3
...as a new installation, not an upgrade.

[Ed_Wilts]

We deployed yesterday. The scheduler failed to properly schedule all the jobs it was supposed to last night.

The call has been escalated within Veritas - it's still a release to be avoided.

[AKopel]

We have MP3 installed on master and all media server along with a Post-MP3 binaries for nbpem and nbjm. This weekend we will find out how it all works....

[Stumpr2]

what are the post MP3 binaries for nbpem and nbjm suppose to fix?

[Lance_Hoskins]

Welcome to the world of NBU 6.0 Bob. Finally we can all gripe together about the same product! :p :) :p

[AKopel]

The post-mp3 binaries are supposed to fix the problems I mentioned in my earlier reply in this thread...

[Alasdair_McQuir]

Has anyone else got this message in eventvwr,


Application popup: nbgenjob.exe - Application Error : The application failed to initialize properly (0xc0000142). Click on OK to terminate the application.

It is somthing I am seeing since installing NBU6

[Stumpr2]

Aaron,
I only understood the earlier post to concern inline tape copy. You mentioned 2 problems but I only see the one. Please tell me what the 2nd problem is. Thanks.
Bob