date recovery 3590E cartrige

[mjvraspir]

Please help, I am confused.



I found a place that claims they can actually recover data from a 3590E (very expensive).

They have my tape and report they "successfully scanned the tape and identified 970 blocks of stored raw data. This data contains a single file called ADSM.BFS.V000F89"

I know that ADSM.BFS is a default data set name indicating the tape will be used by TSM for data storage.



Several questions...

Is that the tape label? (I know I did a label libvol ) this one in my mind is DRA271. We order the tapes 'labeled and initialized' (if anyone has a good handle on this I would appreciate info)

Would the data be stored in a single aggregate file?

Is there a default block size on the 3590E?

I am looking for ~16 G from this particular tape.



Thanks in advance.

Mary V.

 

[n9kws]

Perhaps someone can confirm this, but from my TSM classes I seem to recall the follwoing:



Rule 1) Without the DB, the data on the tape is WORTHLESS, because only the DB has the info where the data for individual files on the tape stop and start.



Chances are the data on the 3590E tape is in a compressed format, so in addition to block size you also have compression to deal with. Compression is completely based upon the source date. For example, my 3590E1A drives LOVE Oracle DB file, and I can get 140-160Gb on one tape easy becasue of the compressing, and sparse file layout of Oracel DB files.



To answer your quesitons:



Q) Is this the tape label?

A) Probably not, if it is a 3590E cart then the library manager is telling TSM the label name. I am told by our purchasing people the labeled and initialized tapes are up to $10 more dollars per tape and the process is so easy that you can do it yourself and save.



Q) Is the data stored in a single aggrate file?

A) See rule #1 above, the short answer is probably YES!



Q) What is the default bolck size?

A) This is the least of your worries, due to rule #1 above.



Hope that helps,

Andy.

 

[chad_small]

Mary,



It looks to me like the volume label is part of the aggregate name. The easy way to tell if that is the label name is to check the last 7 characters and see if they match the vol-sers in your library. TSM by default uses ADSM as the label prefix when setting up sequential device classes. Here is what my 3590 devclass shows for label prefix:



Label Prefix ADSM



So TSM must label the tape and create a single aggregate spanning the entire tape. Like disk volumes if then puts the data into this file. Without the DB I don't know how any company could get the data back. The DB is the only referrence to where the data resides in the aggregate file.



Chad Small

TSM Certified Consultant

IBM Global Services

[email protected]