Volume is allocated to two different servers and storage pools but the volume exists only once

[hanspeterpitz]

Hello,

has anyone experience with the following problem?

Volume A04010 is allocated to two different servers and storage pools but the volume exists only once

17:49:32 TSM2W : q vol A04010
Volume Name Storage Device Estimated Pct Volume
Pool Name Class Name Capacity Util Status
------------------------ ----------- ---------- --------- ----- --------
A04010 TSM2W_J_TA- J43592CLA- 4.2 T 1.3 Full
PE_COPY_RZ SSA

17:48:40 TSM1K : q vol A04010
Volume Name Storage Device Estimated Pct Volume
Pool Name Class Name Capacity Util Status
------------------------ ----------- ---------- --------- ----- --------
A04010 TSM1K_TAPE- J43592CLA- 4.3 T 18.3 Full
_RZ SSA


Thank you for your support.

Best Regards,
Hans-Peter

 

[Harry_Redl]

Hi,
try to audit the volume (audit vol a04010 fix=no) on both servers. If both succeed (process can take a while - monitor "q proc" and "q act" output) you really have two volumes with the same label (can be OK if in different libraries and not shared via Library Manager- although confusing).
If one fails (likely) then the volume is unusable for that server and should be recovered from copy storage pool (if there is any) or deleted.

Hope it helps
Harry

 

[marclant]

Like Harry said.

Two possibilities.

1. You have 2 physical tapes, with the same named, both used on different server.

2. You have 1 physical tape, used on server1 at some point and removed from the library. Checked in a server2, which would be a scratch because server2 is unaware that server1 used it. So now, the tape is defined in both servers. The last server that has written to that tape is the one that has valid data on it. The other server has lost the data on that tape and can potentially be recovered if the data exists in a copy pool.

 

[hanspeterpitz]

Thank you for your advice

I forgot to mention the following

We are in the middle of an upgrade project. Server TSM1K and TSMK2W are old tsm server on TSM 6.3 using (shared) LIB MGR on the new Spectrum Protect 7.3. server.

Best Regards,
Hans-Peter

 

[hanspeterpitz]

Hello Harry,

the audit vol a04010 fix=no results:

Audit Result TSM1K
09/05/2019 14:08:30 ANR2317W Audit Volume found damaged file on volume A04010:

Audit Result TSM2W
09/05/2019 15:31:38 ANR2317W Audit Volume found damaged file on volume A04010:

I guess that's the worst case.

Regards,
Hans-Peter

 

[hanspeterpitz]

Hello,

the same volume is used by two different servers / storage pools and the number of "problem tapes" increases daily, but I can not see why. I am afraid that more new backups will destroy more tapes and we'll have a total loss.

What can be the reason for this problem? Any ideas?

Hans-Peter

 

[Harry_Redl]

Hi,
how is the library in question (containing the "double-owned" tapes) defined on all servers?
Can you show us the output (from all the servers) "q libr <libname> f=d"?
Trying to imagine the way how to get into this state - maybe library defined independently on both servers (therefore not really shared).
And back to original question - can you get us full output of the audit processes? (I am interested in mount messages, library manager/client communication messages and summary output from the audit - numbers found OK, damaged etc.).
Hope it helps

Harry