Hi Experts,

I am looking for your suggestion I am using TSM version 6.1.100 which is interlinked with IBM Library TS3100 from last few days my tapes are filling so early. I checked but did not find anything wrong in the configuration according to my knowledge. the most confusing part is mainly i am using 2 storagepool in one pool whatever the tapes are assign that is not filling so early at the same time in another pool tapes are filling quickly which is not normal I am attaching few snapshot how my backup module is design pls do let me know if any other details which i missed unknowingly.

I have attached snapshot on vmpool whatever the tape I assing it consume the tape quickly. Backup retention policy is same for every node.

Regards

Prem

Prem,

I'd like to help!  This is a question I understand.  One request -- in future, please include output as TEXT (using cut'n'paste) rather than as JPEG screen captures.  This is because:

  (1) You can put the example output with your comments, about the stgpool that works [TEXT], and the stgpool that is broken [TEXT].

  (2) Text is less storage than JPEG, and doesn't require another click, and another window, that blocks the first window.

  (3) When the output is divided into multiple JPEGs or images, you cannot see all the output together, and it is difficult to tell which bottom-half goes with the corresponding top-half of image; it is just a major pain.

  (4) When I want to reference your output, I cannot cut'n'paste; so I need to re-type all your images, or ignore them.

OK, now answering to your question!!

First, you mention the tapes "are filling early" but there are just 2 stgpools in the output, but the VMPOOL is using just 2 tapes, and the SQLDBPOOL is using just 1 tape.  How could you want to use fewer tapes?  That seems almost impossible.  By the way, change MAXSCRATCH on SQLDBPOOL to a much lower number than 9999, because it makes other output misleading.

Do you have many TSM Clients?  Are there lots of filespaces?  How many TB of data? What are the collocation settings?  If you don't know how to get these answers quickly, here are the commands:

  select count(*) as NODE_COUNT from nodes

  select count(*) as FILESPACE_COUNT from filespaces

  select sum(total_mb) as TOTAL_MB_ALL_NODES from auditocc        (after AUDIT LICENSE process completes)

  select stgpool_name, est_capacity_mb, collocate from stgpools order by collocate

If a stgpool is configured to use collocation by FILESPACE, then if you have 6 nodes with 7 filespaces each, TSM will try to put the data on 42 separate tapes, one for each filespace; it will only overlap filespaces on a tape if there aren't enough SCRATCH.

If the stgpool has COLLOCATE set to NODE, then every node gets its own tape (again, assuming enough tapes).

If COLLOCATE=GROUP, then TSM puts all the systems within a COLLOCGROUP onto one tape; so one tape per group.

If you have not change the default, it is COLLOCATE=GROUP.  Which means, if you have not created COLLOCGROUPs, that each NODE is assumed to be its own GROUP -- so you use one tape per node.

Again, you are only using 3 tapes, so there is nothing to complain about.  Also, collocation (by filespace or node or group) is useful when you have multiple backup streams, and it is a huge advantage when you need to do multiple restores!

Hope this answers your questions!

  Best Regards,

      Robert Luehrs

Robert,

Thank you very much for the response and accept my sincere apology for pasting the snapshot as a Jpeg. I will make sure for future transaction i will only paste the text. Pls find the output.

1.tsm: ITSM01>select count(*) as NODE_COUNT from nodes

 NODE_COUNT
-----------
         32

2. tsm: ITSM01>select count(*) as FILESPACE_COUNT from filespaces

FILESPACE_COUNT
---------------
             73

3.tsm: ITSM01>select sum(total_mb) as TOTAL_MB_ALL_NODES from auditocc

TOTAL_MB_ALL_NODES
------------------
         115879199

4. tsm: ITSM01>select stgpool_name, est_capacity_mb, collocate from stgpools order
by collocate

STGPOOL_NAME                EST_CAPACITY_MB     COLLOCATE
------------------     --------------------     ------------------
DISKPOOL_VM                        512000.0
LOTUSDOMINO                      12800000.0     GROUP
MAILPOOL                              196438586.0    GROUP
SQLDBPOOL                     22398577560.8     GROUP
VMDOMINOFINAL           8000000.0            GROUP
VMPOOL                           40744513.1            GROUP

As i told you earlier only 2 stgpool are in use currently other are I use while I have to restore the data of Lotus Notes which is now we are not using in our production.

As you pointed Again, you are only using 3 tapes, so there is nothing to complain about. I am not using only 3 tapes other tapes also assign to storage pool according to requirement but yes 3 tapes are in use.here is the reference for the same.

KG3854L4                  VMPOOL       LTO_CLASS-      1.2 T   40.3    Full
                                        _1
KG3855L4                  VMPOOL       LTO_CLASS-      1.1 T   47.5    Full
                                        _1
KG3856L4                  VMPOOL       LTO_CLASS-      1.1 T   99.9    Full
                                        _1
more...   ( to continue, 'C' to cancel)

KG3857L4                  VMPOOL       LTO_CLASS-      1.6 T   18.3  Filling
                                        _1
KG3858L4                  VMPOOL       LTO_CLASS-      0.0 M    0.0   Empty
                                        _1
KG3859L4                  VMPOOL       LTO_CLASS-      0.0 M    0.0   Empty
                                        _1
KG5061L4                  VMPOOL       LTO_CLASS-      0.0 M    0.0   Empty
                                        _1
KG5062L4                  VMPOOL       LTO_CLASS-      0.0 M    0.0   Empty
                                        _1

One more thing why I don't know few tapes shows full but they have only consume half size of the data space, when i have configure that will contain only 3 version of backup it should clean the previous one and start writing on the same tape new version of data. waiting for your take on this.

Regards

Prem Singh

Prem,

Thank you very much for the cut'n'paste text -- makes life much easier.  Based on the info you provided, I think your TSM Server already has a good configuration, and is working well -- don't see any real problems.  Congratulations!

Since all your stgpools are using COLLOCATE=GROUP, *if* you decide you don't want every NODE in its own group, you would commands like these:

   define collocgroup   group_1  description='Group for first half of systems for VMPOOL'

   define collocgroup   group_2  description='Group for second half of systems for VMPOOL'

   define collocmember  group_1   node_a,node_b,node_c,(...and so on)

   define collocmember  group_2   node_x,node_y,node_z,(...and so on)

Or, you could accept that you have a few more tapes in use -- which again, can be a GOOD THING if you need to do a bunch of restores, assuming you have multiple tape drives to mount the various tapes for each node.  That decision is up to you.

About the percentage being used by each tape, here is a one-page description of reclamation with a cool diagram to explain:

     https://www.ibm.com/support/knowledgecenter/SSGSG7_7.1.1/com.ibm.itsm.srv.doc/c_reclaim_works.html

In summary, it sounds like your TSM Server is working normally.

Hope this helps!

  Best Regards,

      Robert Luehrs

Robert,

Thanks, really appreciate for your time and patience to bearing with me.

Regards

Prem