Media pool help please.....alcohol on offer :)

[myuserid7]

Hi all,

Before I start I'd just like to say that there is a free drink for the expert that helps me with this because its driving me mad....I've been wrestling with this for a while and gave up but its now a MUST to get it sorted.
(Free drink is for the expert deemed to have offered the best advice...not everyone...and i'm the judge )

Ok, here goes,

I have a DELL LTO3 tape library with 2 LTO3 drives.
The unit has space for 24 tapes.
The current backup is 2TB.
The current tape usage for a weekly backup is 4 tapes.

Ok, how do I start to try and get this AUTOPILOT (with GFS I think) and Media Pool stuff sorted out ?
I start by making a media pool and then get hit by :-

Mediapool name :
Base serial number : 1000000
Serial number range : 100000
Min # media in save set :
Renetion time :
Prune retension :

How do I know what to put in these boxes and I know that if I get it wrong i'll be screwed later on down the line.

What happens with the 'Preserve Media' options on the Autopilot options ? (Daily, Weekly, Monthly).

And a whole host of other stuff that I really dont know.

So....can someone basically help me to setup the whole thing and also tell me what reports I need to run to tell me what media needs to be in the library and what tapes need to be in the safe.

Many many thanks for reading...(and hopefully helping)

Scott

 

[blister911]

Scott,

(Regarding the media pools)
Create your backup job and let it create the media pools.

On the Schedule tab, pick Use Rotation Scheme, select the GFS method you would like to use, and then in the Media Pool Name Prefix box put what you would like to have for the names of the media pools. I think it has a max of 7-8 characters. We use the server name (or an abbreviation if its too long).

The backup job will then create 3 media pools for you. If your prefix is SERVER1, it will create:

SERVER1_DLY (daily pool; week days; m-th if 5-day rotation, m-sat if 7-day)
SERVER1_WLY (weekly pool; end of week full; friday if 5-day, sun if 7-day)
SERVER1_MLY (monthly pool; month-end backup; last end of week full backup for the month)

You can change the Min # of media to keep in the save set, as well as the retention time to your own preferences (do it for each pool as you see fit). These setting are for how long you want to keep previous backups before they are overwritten (I think 9999 is the max for both values).

Base serial number and serial number range are for labeling media that don't contain barcode labels. These are set automatically when the pools are created. You can, pretty much, ignore them if you have barcode labels on your tapes.

You will need to go into the media pools and manually move media around to get everything started. After the initial inventory, any formatted media not already assigned to a media pool get put in the DEFAULT media pool. Move required media from the default pool's save set, into the appropriate media pools for the backup job. Only media in the scratch set of the appropriate pool is available for backup.

Good Luck.

 

[myuserid7]

Brilliant, thansk for the top-notch advice. You seem to have put it in a way that is simple to understand (which is what I needed

Just a couple of things:

1. If I fill my library with tapes, will the backup job just happily go along and use the ones it needs to or will I have to remore used media?

2. Which reports do I need to run to tell me (or tell other members of staff I'm not here) which tapes need to be in the library and which ones need to be in the safe.

Thanks

Scott

 

[blister911]

Scott,

1. The backup job will only use formatted media that is in the appropriate media pool scratch set, regardless of how many tapes are in the library. Notice I said FORMATTED. If you have blank, unformatted media in the library, any job (daily, weekly, monthly) can use it. The job will format the media prior to using it and then put it in the save set of the pool that corresponds to the job. If you want to use existing, formatted media you will need to go into the media pools and manually move them.

2. I believe there is a media prediction report that will tell you what media can be used for the next job. It's a good supplement, but I stop using it long ago. Maybe it has changed, but my biggest problem was that it didn't tell you if the media it said could be used for the next backup job was in the library (online) or not (offsite). I still had to go into the media pools and check. So I've just gotten used to checking the media pools and I the only problems I encounter are if I forget its a monthly backup and check the weekly pool for media.

(Just my opinion here). Train your helpers on how to use the BA Backup Manager. Try to avoid the "trained monkey" syndrome. What I mean is, you can train a monkey to change the tapes, however, there will come a time when your helpers will have to think for themselves; like when the media prediction report tells them only blank media can be used for the next job.

We use 5-day differential backups, so our daily backups are 1 tape (8 tapes total). We also have 24 slot LTO3 libraries. That leaves 14 slots in the magazine for weekly (one of the five sets are always off-site) and monthly backup tapes (only one set is ever online). One of the daily tapes is always off-site and 2 slots are reserved for cleaning tapes, plus the mail slot….I think that comes our right.

That will not work as well for you, but it gives you an idea. Twenty-three magazine slots may not be enough to hold all your backups.

The key to getting all this to work properly, is setting the right values for Retention Time (Rt) and Min # Media In Save Set (min#). I've heard several different opinions on how to set these. You will have to figure out what works for you. Someone else may be able to offer you a better way to set these up, but here is what we are doing:
server1_dly min#=7; Rt=13 days
server1_mly min#9999; Rt=9999 days (we never overwrite these; saved almost for forever)
server1_wly min# 4; Rt=34 days

Note: For backup jobs that traverse multiple tapes, the min# is regarding the sets not the individual tapes.

Good Luck

 

[myuserid7]

Excellent, going to run with it this weekend.

So....how do I get this drink to you then ?

 

[blister911]

No need. If I was in it for the "fringe benefits" I wouldn't be posting here.

Good Luck, I hope everything works out well for you this weekend.

 

[myuserid7]

I don't flippin (you know the real word here) believe it !!

It started out out ok, then my job crashed and I think I now have a corrupt database.
Why oh why does this keep happening...This will be the third time in a few months that I've had to start a new database. Is this normal ?, should my database directory grow so large after such a short period of time ?, is there anything I can do to sort this out for good ?

So it looks like I've cracked the autopilot stuff but now my Brightstor is up the wall....brilliant.

Thanks for all your help.

Scott

 

[blister911]

What version of BrightStor are you using? Usually, the first piece of advice is to apply the latest service pack if you are not already running it.

Are you getting any error messages?

 

[myuserid7]

Its Brightstor 11.5 with SP2.

The error messages just refer to the fact that it can't write anything to the database - and the database engine then stops. I have created a totally new database and it is running ok but what measures do you think I should use to make sure that my database stays at a nice size (I think this is the trouble I get every few months with my database growing too big).

Regards

Scott

 

[blister911]

Scott,

Go into Server Admin (via program menu or the manager). Change the server to the one you want to administer ("Select Server" icon or do Admin\Select BrightStor Arcserve Backup Server). Then click the "Config" icon or do Admin\Configuration. Switch to the Database Engine tab and set/change the prune database settings. I think by default it's set to prune records older than 999 days. Click the help button at the bottom to read up on what the various parameters do.

I think that should keep your databases smaller. We are mostly NetWare, so our config is a little different. But, we have ours set for 90 days. Any restore request for files older than 90 days we have to run a merge before we can do the restore.

If you are having issues every couple of months then I would set the value to something under 60 days.

I hope that helps you some.

 

[crispychimp]

In regards to your DB corruption issue - I had the same problem when I started with Brightstor.

It sounds like no database maintenance is being run on your db. If you are using the ARCserve native db, then see the instructions below:

http://supportconnect.ca.com/sc/kb/...&docid=389855&bypass=yes&fromscreen=kbresults

You can use the Windows Scheduler to automate the routine.

If you are using SQL Server, you may want to check out your server config, because it should be maintaining itself.

 

[olliday]

Hi,

About your DB issue :

You state BAB11.5 SP2, I'll assume build 4232.

1) crispychimp is right, let me add that you'll find a sample maintenance script here :

http://supportconnect.ca.com/sc/kb/...&docid=268789&bypass=yes&fromscreen=kbresults

For mssql, set a maintenance plan, cf :

http://supportconnect.ca.com/sc/kb/...&docid=394425&bypass=yes&fromscreen=kbresults

2) Either apply :
- QO84317 NT -HOT BACKUP CAUSES VLDB TO BECOME CORRUPT

http://supportconnect.ca.com/sc/sol...?aparno=QO84317&os=NT&returninput=0&prev=true

Or
- QO86307 NT -SERVICE PACK 3 FOR BAB R11.5 FOR WINDOWS

http://supportconnect.ca.com/sc/sol...?aparno=QO86307&os=NT&returninput=0&prev=true

3) If you run an antivirus on BAB server, exclude BAB\database folder from realtime scanner.

4) Then start with an empty, clean DB

5) If you still have DB issues, consider switching to mssql.

http://supportconnect.ca.com/sc/kb/...&docid=390902&bypass=yes&fromscreen=kbresults

Regards.