Usual reasons were already explained but there is another one I had
recently:
A node is using client compression with large number of threads (16-20).
Each thread is allocating some space in disk pool and starts
reading+compressing the data. If space is very close to *compressed*
ammount of data to be stored few last started threads attemted tape
writes. Space was not enough to be allocated and first threads were not
finished compressing to reduce allocation with actual size. From there
came some tape mount requests.
The case was:
- 600 GB Oracle DB spanning more than 20 filesystems
- data files of 1,2 or 4 GB
- compressed usage was about 130-150 GB
If we had less than 160-170 GB in disk pool we've seen occasional tape
mount requests close to end of backup. Compression ratio highly depends on
data contents and we cannot blame TSM for being unable to predict
precisely.

Zlatko Krastev
IT Consultant




Please respond to "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
Sent by:        "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
To:     ADSM-L AT VM.MARIST DOT EDU
cc:

Subject:        ANS1312E  Server media mount not possible

I assume that "ANS1312E  Server media mount not possible

        Explanation:  Server media mount not possible. The server timed
out
        waiting for a mount of an offline volume." refers to mounting a
sequential acess media?  If my assumption is correct, under what
conditions should a backup process request a server media mount?

I am getting ANS1312E for a node which goes to a disk storage pool
which has a max size threshold of NOLIMIT.  I would not expect it ever
to request a server media mount.

Any ideas?

David EHresman