2010 Media Sets - Pulling from other sets when the tape is not full?

[Theydon_Bois]

I have jobs set up to specifically target a dated Media set, Monday1 through to Friday6.

A job from the server with an attached Superloader3 Ultrium 3 device backs up under 300Gb to the specic tape, the job running under 8hrs.

Another job from a server running remote agent then runs, backing up some 10Gb or so to the same media set appending to the same tape.

This should give me the contents of two servers to one tape.

This all worked fine for a period of time, and then I started getting problems, with it pulling media from other tapes despite requesting a speicifc media set, and now the 300Gb job is backing up partially to one tape, pulling another tape to continue and then the other tape backs up another tape it feels like.

For example:

Monday: Server A (300Gb) ignores Mon4 (this week) and pullsTues4 and Wed4. Server B (10Gb) pulls the Mon4 tape as it should.

Tuesday: Server A pulls Tues4 and Wed 5 and Server B pulls Wed5.

So why is a 300Gb job pulling 2 Ultrium 3 400Gb tapes for its jobs? What is the piont of specifying a media set if the server ignores you and pulls tapes from other sets? tbh, I would rather the job fall than overwrite other jobs or cross multiple tapes and potentially ruin other good backups.

I have tried various different settings, and my jobs are set up as Server A: Device and Media: Overwrite Media, Server B: Device and Media: Append to Media and the Media Sets are set as OPP: 0 Hours and AP: Infinite, Allow Append.

I understand that the media sets and my configuration allow others to over ride this, but this should only happen if the tape fills (and I will ensure that this will not happen) but it must be set like this to allow Server A and Remote Server B to backup to the same tape.

Help!

[AmolB]

It is always recommended to keep the tapes in the scratch media. Backup Exec will pull the

available tapes from the scratch media and then once the backup is done the tape will be

in the media set associated with the backup job. If the job is spanning over multiple tapes

then BE will always look for an overwritable media.

[Kiran_Bandi]

As you said, for some days backups worked fine as you expected i believe you are not moving media manually into the media sets. If not, as Amol already said it is not recommended to do that. As you configured your first servers job to overwrite, if you move media manually into the media set also it should work. What is the media management setting you configured? It should be overwrite media in the target media set before overwriting scratch media.(Select from Tools--> Options --> Meida Management). And select Overwrite protection level as full. The combination of above two settings make BE to overwrite media in the following order: Recyclable media in the targetted media set, Scratch Media, Recyclable media in any media set. Try these settings and provide update here.

Regards...

[Theydon_Bois]

Kiran, both of the settings you mentioned were not set as you indicated.

I have changed them both now so that it "...Overwrites to the same media set" and Overwrite Protection is now set to Full.

For reference, I have 30 Media sets created, one per day Monday1 to Friday6, containing only the tape relevant to that day. Scratch media contains no tapes.

Amol, the reason why I used Media Sets instead of just scratch media was to try and ensure the two servers backed up to the same day. Otherwise Server A will backup to Mon5 and ServerB to Wed4, then during the course of the 3 weeks, the servers will overwrite or apend to different tapes and be a nightmare to recover from later down the line. Plus rather than one EOM tape, I have two.

I am off now for a few days, so when I return on Tuesday, I will be able to see if these changes have made a difference. I also decided to erase the data on the three days (Thur4, Fri4, Mon5) that I am off in case of append/overwrite issues - one tape has decided to get stuck though, so just sorting that out now!

Thanks again.

[Jayb]

Hi Theydon,

You might have to look at the followng article.

Article URL http://www.symantec.com/docs/TECH5459

And the order of picking is also based on the allocation time , for which the library should be set to  random mode.

Other alternative is creating Partition.

Article URL http://www.symantec.com/docs/TECH27909

[pkh]

 why is a 300Gb job pulling 2 Ultrium 3 400Gb tapes for its jobs?

It could be that you are backing up uncompressible data and getting a bigger resultant file than the original.  See an explanation in my article below.

https://www-secure.symantec.com/connect/articles/compression-short-explanation

What is the piont of specifying a media set if the server ignores you and pulls tapes from other sets?

When BE needs tapes it will first use what is available in the target media set, then the scratch media set and lastly another other media set in the library.  This is by design so that the job don't fail when there is an available tape.

It is actually easier to let BE and the tape library manage the tapes for you rather than you managing the tapes yourself.  Read my article below

https://www-secure.symantec.com/connect/articles/must-mondays-backup-be-monday-tape