Restoring Database to new HW Config

[hjeckman]

I have a TSM Server, SAN and Tape library at one end of the state that is our primary site. I have a server w/ TSM installed, SAN, and tape library at the other end of the state at my DR site.
I'm planning to do a DR test in the near future.
My question (this may sound quite general and stupid) is what steps will I need to perform to restore a working TSM environment to my DR site?
The reason I ask is I have documentation to restore the TSM DB, volhist, policies, etc. to the primary site, but I don't think the same procedures will work at the DR site because the WWN's, paths, etc of the hardware are different.
Thanks for any help that can be offered.

 

[hjeckman]

Hmm..
I may be somewhat ahead of the game then. I have my database backup copied to a share on my DR TSM Server everytime the dbbackup is run, so I won't need to do the db restore from the tape drive.
I was worried about the new TSM Server w/ the existing DB information not recognizing any of my hardware once I have it restored.
If I'm understanding you correctly I should be able to restore the TSM DB, delete all of my devices and then add what I've got out there.
My existing devcnfg at my DR site already has all of the hardware that exists there associated in it.
Or am I overcomplicating this process w/ what I already have set up?

 

[hjeckman]

Ok, I need to elaborate on this, and would appreciate some patience from you guys.
I have my TSM Server at my main site. Daily, I do a DBBACKUP that goes directly to the TSM Server at my DR site.
I also do a drplan file creation that is copied to my DR TSM Server.
I'll use the planexpl.vbs to explode this file into about 25 instruction/.bat files.

The TSM Server at my DR Site already has it's tapedrive and SAN associated w/ the TSM Installation to mirror what I have at my main site.

What do I need to do to complete a DR test just for the TSM Server at my DR Site, you know, restore my configuration to that server. If there's an instruction somewhere you can just point me to that, but I'm killing myself trying to wrap my brain around what needs to be done here.
If there's further info you need, I can provide that as well.

Thanks so much in advance.

 

[hjeckman]

Bump..
Anyone???

 

[d.veeresh]

Ok, I need to elaborate on this, and would appreciate some patience from you guys.
I have my TSM Server at my main site. Daily, I do a DBBACKUP that goes directly to the TSM Server at my DR site.
I also do a drplan file creation that is copied to my DR TSM Server.
I'll use the planexpl.vbs to explode this file into about 25 instruction/.bat files.

The TSM Server at my DR Site already has it's tapedrive and SAN associated w/ the TSM Installation to mirror what I have at my main site.

What do I need to do to complete a DR test just for the TSM Server at my DR Site, you know, restore my configuration to that server. If there's an instruction somewhere you can just point me to that, but I'm killing myself trying to wrap my brain around what needs to be done here.
If there's further info you need, I can provide that as well.

Thanks so much in advance.

Break the prepair plan. Then store db, devconfig, volhist and log. device are different on test site. Update device & path .stop shedules then start the restoration part.

 

[hjeckman]

Break the prepair plan. Then store db, devconfig, volhist and log. device are different on test site. Update device & path .stop shedules then start the restoration part.

Couldn't I use the devconfig that's already on the DR server since the tapedrive and SAN are already associated w/ the TSM installation?

And what exactly do you mean by "start the restoration part"

 

[hjeckman]

Bump...
Anyone?

 

[d.veeresh]

I have a TSM Server, SAN and Tape library at one end of the state that is our primary site. I have a server w/ TSM installed, SAN, and tape library at the other end of the state at my DR site.
I'm planning to do a DR test in the near future.
My question (this may sound quite general and stupid) is what steps will I need to perform to restore a working TSM environment to my DR site?
The reason I ask is I have documentation to restore the TSM DB, volhist, policies, etc. to the primary site, but I don't think the same procedures will work at the DR site because the WWN's, paths, etc of the hardware are different.
Thanks for any help that can be offered.

If it is different device on DRsite then redefine drives & paths.Otherwise update in plan file.

 

[mikeatkc]

Couldn't I use the devconfig that's already on the DR server since the tapedrive and SAN are already associated w/ the TSM installation?

And what exactly do you mean by "start the restoration part"

I think you pretty much got the process down. Now just execute and see what happen. TSM DR plans will tell you what you need, i.e. what volumes, to restore your data. The DB, volhist, devconfig, etc. contains enough information to bring TSM into a functional environment. You still need your data.

Good way to do this is perform DR on your DR site and see what steps are done properly and what additional steps you need, etc.

 

[hjeckman]

I think you pretty much got the process down. Now just execute and see what happen. TSM DR plans will tell you what you need, i.e. what volumes, to restore your data. The DB, volhist, devconfig, etc. contains enough information to bring TSM into a functional environment. You still need your data.

Good way to do this is perform DR on your DR site and see what steps are done properly and what additional steps you need, etc.

We are going to the DR site at the end of July to give it a shot and document what works and what doesn't so that we're ready in the event of a true DR event.. I'd just like to have as much ready and planned as possible when we get started.

So, would the batch files that are broken out of the DRplan file take care of everything if I execute them?
My problem is that I've got this great setup and noone in our office has been able to get any real TSM DR training so we're stuck trying to piece together what we can find in documentation, and from peers on the web.

 

[dms19]

Man...with that set up, I'd just mirror the logs and DB at the DR site in the event you lose the main. If the case of distaster all you'll have to do is "flip the switch" and start using the warm site.

 

[d.veeresh]

Man...with that set up, I'd just mirror the logs and DB at the DR site in the event you lose the main. If the case of distaster all you'll have to do is "flip the switch" and start using the warm site.

Good suggestion.

 

[hjeckman]

Man...with that set up, I'd just mirror the logs and DB at the DR site in the event you lose the main. If the case of distaster all you'll have to do is "flip the switch" and start using the warm site.

Now, that's the kind of response I was hoping to hear! We put all this hw and effort into having the most comprehensive DR site and I was hoping in some way it would make this easy.
My DB is already exported to the DR server daily, so one part is already done... Which logs need to be exported?
I'll have my current policy already imported as well.
Will I need to load the dsmserv.dsk or dsmserv.opt files? I'm assuming..
My devcnfg should be good to go already.

Is there any special procedure to restore the TSM DB or do I just need to put it in the location that the server is looking for it?

And then I would just need to have my clients look for the new server name in the dsm.opt file I would think?

Sorry for all of the rookie questions, as I said noone in my shop has had any actual training on the DR portion of the software. IBM came in and set this up and the people involved in that are all gone.

 

[hjeckman]

Bump...
Anyone?

 

[hjeckman]

Sorry to be a PITA, just trying to get answers to my last post.
Thanks.

 

[dms19]

Now, that's the kind of response I was hoping to hear! We put all this hw and effort into having the most comprehensive DR site and I was hoping in some way it would make this easy.
My DB is already exported to the DR server daily, so one part is already done... Which logs need to be exported?
I'll have my current policy already imported as well.
Will I need to load the dsmserv.dsk or dsmserv.opt files? I'm assuming..
My devcnfg should be good to go already.

Is there any special procedure to restore the TSM DB or do I just need to put it in the location that the server is looking for it?

And then I would just need to have my clients look for the new server name in the dsm.opt file I would think?

Sorry for all of the rookie questions, as I said noone in my shop has had any actual training on the DR portion of the software. IBM came in and set this up and the people involved in that are all gone.

You'll need the vitals for sure (dsmserv.opt, devconfig and volhist - that requirement never changes). In the event you had a "fail over" you'd have to do a little tweaking to fit new hardware, IP's, etc.

If the DB is replicated daily you'll have recovery from that point. Provided you have the db, log and stg volumes defined and available on the DR server.

Depending on the type of disaster, if possible having both the primary and recovery servers seeing the same disk (SAN) for db and log volumes protected again by disk replication it could be as easy as "flipping the switch." Plus you won't really see a loss of time or data.

If all the replication is built in, you'd have to change IP's on the TSM server (along with some config parameters) and update DNS, that is if you have dsm.opt/sys files set to DNS instead of TSM server IP and you're good. Otherwise you'd have to update each client to see the new tcpserveraddress of the DR TSM server, which is why we use DNS name.

 

[hjeckman]

Thank you!!
Last question, I believe.
I'm assuming the dsmserv, volhist and devconfig I can just stop the server and put those in place (replace the existing) and then what do I have to do w/ the DB? It's exported to a file share that's seen as sequential media by TSM. Would I just need to copy it to the drive that the DB resides in, essentially replacing the skeleton DB that's there and start the server?

Also, how do I know what DR Tape volumes I'll need for restores?

Will running XXXXX.RECOVERY.SCRIPT.DISASTER.RECOVERY.MODE.CMD
perform all of these steps that I'm asking about? If you're not sure what file I'm talking about I can copy and paste the commands it executes here.
I've gone through the dr plan that get exploded and that's the only one I see that I have to run, the other ones look like I would use them if brining the original server back up from a crash or something.

I know what to do w/ the clients once I have the server up, my only hang up is w/ the TSM Server itself.

Thanks again, you all are great.

 

[dms19]

Best bet in this situation is to restore the database. Just handle it like a normal DR situation. If you have the vitals, you restore the database and go from there. The volhist will contain all of the volumes necessary for recovery if accompanied by the latest dbb. You'll have to have the copy group tapes to be checked into the library and do a little tweaking there if you also need to restore client data, but for functionality you'll just need to have the db/log and storage pools in place and once the restore of the dbb is complete its more or less ready to go.

 

[mikeatkc]

Thank you!!
Last question, I believe.
I'm assuming the dsmserv, volhist and devconfig I can just stop the server and put those in place (replace the existing) and then what do I have to do w/ the DB? It's exported to a file share that's seen as sequential media by TSM. Would I just need to copy it to the drive that the DB resides in, essentially replacing the skeleton DB that's there and start the server?

Also, how do I know what DR Tape volumes I'll need for restores?

Will running XXXXX.RECOVERY.SCRIPT.DISASTER.RECOVERY.MODE.CMD
perform all of these steps that I'm asking about? If you're not sure what file I'm talking about I can copy and paste the commands it executes here.
I've gone through the dr plan that get exploded and that's the only one I see that I have to run, the other ones look like I would use them if brining the original server back up from a crash or something.

I know what to do w/ the clients once I have the server up, my only hang up is w/ the TSM Server itself.

Thanks again, you all are great.

The TSM disaster recovery file should indicates what volumes go where and you can view the listing. After DB restore, TSM will let you know what volumes are required (if missing), monitor the activity log.

 

[hjeckman]

Thank you all so much!
Which file lists the volumes needed, is it XXXXX.COPYSTGPOOL.VOLUMES.AVAILABLE.MAC? When I break up the drplan file there's about 25 different files.


I think I'm just about ready for a hands-on crashcourse in TSM DR now.
Wish me luck.