Suggestion needed to Speed up Retrievals from TSM

[jneetz]

Issue:  Running multiple ARSDOC GET and even if just running one it is averaging 3-6 seconds a document sometimes less
      ( mix of PDF, TEST. IMAGES )

System:  On LINUX Server older CMOD 8.5 and using Tivoli on the same server

We already upgraded to the latest CMOD Version ( and Tivoli ) and new Storage Hardware
    However still going through the Migration of Data  ( 3+ TB and millions of Documents )

Process:   Running ARSDOC GET by AppGroup and Date Ranges ( Depending on Groups Range can be Week or Monthly )
               Then process the output ( IND / OUT files ) into the New System

It's all straight forward and working .. Just very slow for the amount of Data we have
     Running multiple ARSDOC helps some but then as assumed slows each retrieval

Bottle neck seems to be something with TSM and possible the DB2SYSC using high CPU on the LINUX server

ANY SUGGESTIONS
    Can TSM be tweaked somehow, maybe using different DB2SYSC processes for each ARSDOC get somehow
    ** I'm just thinking as typing, maybe it's not db2sysc as the issue

Possibly how can a Whole AppGroup ( and maybe by date range ) be sent to Cache Storage before I do retrievals against it
    Or will that really save time sense it's kinda a double get

thanks

[jsquizz]

I hate to bump such an old OLD thread.. However I think you're still an active user.

What did you decide on?

[jneetz]

Sorry just seeing the reply from March

I am running a SQL against the AppGroup Table in the database ( Only 6 AppGroups )
   This creates an Index List ( doing to for one month at a time )

I have a vbsscript which then creates a PARMFILE to be used by ARSDOC -F option

Thing is though I then had to add the ARSDOC parameters of, else it kept stopping on me
    -n retrieve items one at a time from the server
    -s 1  waits between query requests

Still somewhat slow, or maybe as fast as it can be
    But is consistent and not randomly stopping