Search:

This is not particularly accurate but it can give you a sense of which tapes have been reused most. It is based on the assumptions that you have an extensive volhistory and each pair of stgnew and...

I am not an expert with Powershell or windows scripting but have you tried just single quotes?

.\dsmadmc.exe -id=admin -password=xyz123 define schedule Admin01 type=administrative cmd='query...

You can specify a particular archive if you know the date or if you supplied a description as part of the "dsmc archive" command. Use the "query archive" client command to see what is available.
...

The command you give is to retrieve an archived file, not to restore a backed up file. They are not the same concept. If the file was backed up (via dsmc incr or dsmc sel) use the dsmc restore...

Your current TSM instance will not be able to access those files. Assuming you have a DBB from before the expiration you could use it to restore your TSM instance to that point in time and then...

Earlier in the thread didn't you say you have 2 tape drives?

Look at the RECLAIMSTGPOOL option for storage pools. Define one for your tape pool, make it devc=FILE and then make the NEXTSTGPOOL of...

Is the client only backing up 2 file systems or 2 drives? If so then you will not get more than two mounts as TSM will not "stripe" the backup of a filespace across multiple mounts.

Actually, assuming that all of your TSM servers use your configuration manager to manage admin accounts then you should be able to delete the account on the configuration manager and then issue a...

You have collocation=group, do you have collocation groups and are your nodes assigned to those collocation groups? Another item to check is the number of migration processes you are running...

No. PCT_Util refers to the client filespace utilization, in this case /DATA is 19,922,944 MB in size and it is 94% filled. Which means a theoretical full backup of that space would send something...

I am not aware of any shortcuts. Look for the dsmserv.opt files.

Why did you clone the instance instead of halting the original, restoring it onto the new HW and then promoting it to a library manager?

A query like this will tell you all of the primary tapes that "might" be needed to create a backupset

select distinct volume_name from volumeusage where node_name='<NODE_NAME>' and...

It seems that you would have to do the following on Mondays:
- perform the daily eject job to the local Iron Mtn
- perform the second job but this time specify a tolocation in the drm command...

If you mean the value of the "Vault Site Name" as seen in the query drmstatus output, no there can only be the one. However, the move drm command allows the addition of WHERELOCATION and TOLOCATION...

Ok, how about you update the volume to scratch (update libv ... status=scratch) and then try a label libv on it? This will not work if TSM believes the tape to be already in use as a stgpool volume...

jbiggers,

We do have server groups and library managers and the tape movement between two different library mangers was the same as with movement between two libraries managed by the same library...

jbiggers,

We have on a number of occasions moved thousands of tapes between libraries. And we also used one library, with sufficient empty slots, to label new and used tapes with new barcodes for...

Did you perform a "checkout libv ... checklabel=no" command?

Patravane,

Have you tried checking in one of the suspect tapes using the mail slot and with checklabel=yes? Did you get any errors? Do you have copy pools?

Do you have a management class called "dir_class" in the domain and active policyset the new nodes are using?

That should be all you need to do provided that all of the management classes the node used in the original domain/policyset exist in the new domain/policyset. If management classes are missing then...

Assuming you have a SCSI library and your problem is you only have 1 I/O slot, you can change your first command to

.... remove=bulk .....

which according to the documentation (help move...

The message states there is an extra parenthesis in the command. That section should resemble this, and note I am adding spaces around the parentheses to avoid emoticons.

cast( ( verexists ) as...

You can but I think it would look ugly in cases where there are multiple classes in the active policy set. The schedule would appear on each line. And of course with every additional join the query...

You probably also want to add

.... order by n.node_name,n.domain_name,Set,MgmtClass

to the end of the select in order to group the nodes together. This select, with the suggested...

kenshoomy,

I did not quite follow the reasoning that got you to a max of 64 partially filled tapes. Although by your example it seems like each NAS uses multiple threads. I had assumed that with...

So right now you have a group collocated pool but no collocation groups, so effectively you are node collocated. You need to keep in mind that group collocation with only one group means that data...

I don't have any TSM 6.x experience but I can offer something that I think could work in 6.x, albeit it is extremely ugly and convoluted, and perhaps not even worth the effort. It was tested out on...

To my knowledge there is no way to see what TSM is doing during this kind of query. I have a few really large DB's (V5.5) on which I can't even do this kind of query without it eventually dying. ...

If you are using the "LIKE" operator you need wildcards. Try this

select node_name,hl_name,ll_name,backup_date from backups where node_name='REGLVFILE1_GROUP' and hl_name LIKE...

We have an admin schedule that runs a script. In the script we use the parallel and serial commands to group threads together. Example

Schedule Definition

Schedule Name:...

Something like this might be what you want. Replace the list of pool names in () with your own primary pools

select node_name,cast(sum(physical_mb)/1024 as decimal(10,2)) from occupancy where...

I periodically get a similar situation with my dsmserv processes. We disable the /etc/inittab line to prevent auto start of TSM on boot and instead use a start-up script that is also used by our HA...

I think you could use multiple dsmcad daemons where each one specified a different option file using the -optfile command line option. I used to do this with multiple Domino clients on the same...

Mikey D asked earlier if the owner/group of the volumes was the same as the account running the dsmserv binary. You replied by listing out the volumes on both systems. So I think the question still...

TSM scripts are very limited but you can do the following with your example (days of week start with Sunday as 1)

select server_name from status where dayofweek(current_timestamp) in (7,1)...

You can't specify a file but you can use a comma delimited list of nodes.

q eve * * node=NODE1,NODE2,NODE3,...,NODEN ....

Spiffy,

Tapes go to vaultretrieve after they go pending and then meet their storage pool reuse delay, either through the data expiring or through the reclamation process where replacement copypool...

Like this. Note I had to guess on the char() for status since I see an emoticon. I also used the "in" function near the end.

select cast((server_name) as char(10)) AS TSM_SERVER, \...

You wcan't just compare a timestamp to a date, you need to use a function to convert the timestamp.

date(scheduled_start)=current_date

yes, current_date

I see when you run it via the console you are "root". When you run it as a script is it also root and/or does it have the same TSM environment variable values as when you run it manually? If not...

If you look at the output of the query occupancy command for a node you will see the amount of data that each filespace has in each storage pool. So you can issue a variation of the select below to...

You can query the summary table using a select similar to this

select cast(sum(bytes)/1024/1024/1024 as decimal(9,3)) as "GB" from summary where entity='NODE_NAME' and start_time>'2012-07-25...