onsite & offsite pools not synced

[rcrowder]

For some reason, my offsite copy pool is not in sync with my onsite pool. I have tried using the backup stg onsite offsite but it completes with no copied files and no errors but when you compare the files that exist in the onsite and offsite pools they do not match. I know this works because I have another TSM server and its fine. Somehow, I think that some of the data got put (migrated) on the offsite pool and not the onsite pool or something. Is there any way to audit or run a compare on the two pools to get them synced??



Thanks in advance,



Ron

 

[run]

TSM shouldn't allow you to migrate data directly from clients (without putting the original in a primary storage pool) or migrate directly from a primary storage pool to a copy storage pool.



when running the backup stg, did you try the wait=yes flag so that you can see all the messeges.



Otherwise, what tools/commands are you doing to verify your conclusions that

that files in the copy storage pool are not the same ones in the primary storage pool??

 

[rcrowder]

Thanks for the reply. The clients backup to a primary disk pool. We then backup the pool to tape (offsite copy) then migrate the disk to tape (onsite primary). I'm not for sure what's going on. Here is how I check both tsm servers. This has worked pretty well in the past. The larger tsm server still gets synced everyday.



select sum(num_files) as "Total Tape Files" from occupancy

where stgpool_name = 'TAPE'

select sum(num_files) as "Total Off. Files" from occupancy

where stgpool_name = 'OFFSITE'



When I run this script and the numbers do not match, I issue the backup stg onsite to offsite to sync the pools back up. This has worked for 4+ years on our TSM server. Someone else has been administering the smaller tsm server for a few years now (sister company) and I'm not for sure what is going on. It worked after I installed the server and set everything up for them. Now I cannot get the pools to sync.



Thanks

 

[cartman]

HI,

I have just run those select commands on my TSM server , mine dont match either ... maybe this is normal. Im running TSM 5.2.1

Cartman

 

[okolsteren]

Hi all,



I have the same problems.



Local, I have 2 Tb data stored.



Remote, TSM thinks I have 10 Tb on the local site !!!!



Does it have something to do with the fact that the data is stored as archive on the remotesite. What does the option "Retain versions" mean in this archive copy pool.



If I set this option on (for example) 60, will that mean that all the data on the remote site will be deleted after 60 days and there will be a new copytransaction from the localsite to the remotesite again ??



I hope someone can help me with this one, because I don't want to start all over agian..



Gr. Oscar Kolsteren

 

[rcrowder]

The amount of data may not match exactly because of compression or what but the number of files should as long as you have the same type of file in each pool you are checking. By this I mean, you can't include archive data in one pool and not the other. Our archive data goes to a completely different pool. You also cannot check the pools if you are running any processes that are currently changing the data (backups, migration...). I'm still checking in to the situation.



Thanks for the replys.

 

[run]

Ok, I'm on TSM 5.1.6.5 (AIX) and I ran the 3 sql commands to see what I got. these are slightly different than the ones below: (DR-POOL is the name of my only copy storage pool)



select sum(num_files) as "Total Tape Files" from occupancy where stgpool_name <> 'DR-POOL'



select sum(num_files) as "Total Tape Files" from occupancy where stgpool_name = 'DR-POOL'



select stgpool_name,sum(num_files) as "Total Tape Files" from occupancy group by stgpool_name



(On this last sql statement, I manually added these up and compared them to the DR-POOL. This way I didn't count any storage pools that I was not backup up to offsite)



Since pretty much all my storage pool backup to the dr-pool, I compared the first two sql statement. For the most part, the numbers match,expect the dr-pool seemed to have 4 more files than the onsite pools. So, looks like my machine/config is ok.



You might give the above a go and especially with the 3rd one you might be able to subtract and find that the difference might match some storage pool.



good luck and may the swartz be with you

 

[okolsteren]

Hi rcrowder,



I didn't mean that we backup our archive data to off-site storage, but the documentation says that "At the target server, the virtual volumes from the source server are seen as archive data"



(Administrator Guide - Using Virtual Volumes To store Data On Another Sever)



So, my question stands, what should I fill in in the option "Retain Verion" ??



Gr. Oscar

 

[rcrowder]

run,



Here is the output of my script and the third select statement you gave. I'm not disagreeing with anything you say, I'm just trying to figure out why I can get one tsm server to sync the pools but the other one I cannot. I'm still not getting this. I would have to believe after all processing (expir, backups, migration) that the two pools would have the same number of files. Maybe I'm wrong but it has always worked on my server. When they are not in sync, I would have to wonder which pool has the correct files. I know if there are any tapes that tsm cannot read from it would be possible that they wouldn't show the same because the backup wouldn't finish cleanly, but other than that??? Hopefully I will figure it out... Thanks for your help.



tsm: TSM1>run check_status



Total Tape Files Total Tape MB

---------------- ---------------------------------

39026533 17042937.17



Total Off. Files Total Offsite MB

---------------- ---------------------------------

39026533 17042937.02



Tape Volumes

------------

574



Offsite Vol.

------------

587



Total Volumes

-------------

0

ANR1462I RUN: Command script CHECK_STATUS completed successfully.







tsm: TSM1>select stgpool_name, sum(num_files) from occupancy group by stgpool_name



STGPOOL_NAME Unnamed[2]

------------------ -----------

ARCHIVE 3257903

ARCHIVEPOOL 69632

OFFSITE 39026533

PRODSTAGE1 7

PRODSTAGE2 3944

TAPE 39026533



tsm: TSM1>

 

[rcrowder]

okolsteren,



Sorry about not replying earlier, let me check a few things when I get back to the office tomorrow and see if I can find anything.



Thanks

 

[rcrowder]

After looking at this, I have to say that I have never used virtual volumes. I wouldn't want to post or suggest something without using it. Under normal archives, yes the retain version would be the number of days the archive would be stored. After expiration, the archive would drop off and if another archive was performed, that one would be there for the "retain version" number of days. I would try and start another thread and see if anyone uses virtual volumes.



Thanks



<TABLE BORDER=0 ALIGN=CENTER WIDTH=85%><TR><TD><font class="pn-sub">Quote:</font><HR></TD></TR><TR><TD><FONT class="pn-sub"><BLOCKQUOTE>Hi rcrowder,



I didn't mean that we backup our archive data to off-site storage, but the documentation says that "At the target server, the virtual volumes from the source server are seen as archive data"



(Administrator Guide - Using Virtual Volumes To store Data On Another Sever)



So, my question stands, what should I fill in in the option "Retain Verion" ??



Gr. Oscar</BLOCKQUOTE></FONT></TD></TR><TR><TD><HR></TD></TR></TABLE>

 

[rcrowder]

I don't know if anyone is interested or not that has followed this thread but I used this select statement to find out what files where out of snyc between my pools. May come in handy later.



select filespace_name as " Filespace name ", sum(num_files) as "# of files

" from occupancy where stgpool_name='TAPE' group by filespace_name > filespace_tape.txt



I ran this once for onsite and once for offsite. Imported the files into excel and compared the numbers to find out what was out of sync.



Thanks

 

[AndyRH]

Has anyone found that there are objects in the Copy pool that are not in the on-site pools? We have a reclamation process that fails due to unavaialbe data. We did the previously listed queries and found about 130,000 more objects in the copy pool than are in all of the local pools.

Any ideas how to fix this?

What causes this?



Andy