Some suggestions:
SETSYS DUMPIO(4,4) --> Read a cylinder at a time when backing up or
migrating data sets
PATCH .MCVT.+391 X'05' VERIFY(.MCVT.+391 X'00') --> INCREASE THE
NUMBER OF I/O BUFFERS USED FOR A DATA SET TO 5 WHEN IT IS BACKED UP,
MIGRATED, RECALLED, OR RECOVERED (DEFAULT IS 1)
PATCH .BGCB.+2E X'003C' --> AUTOBACKUP ENQUEUES THE NEXT DATA SET TO
BACKUP BEFORE FINISHING THE CURRENT DATA
SET, BUT ONLY IF THE CURRENT DATA SET IS SMALLER THAN THIS VALUE X 100
(OA10220)
If you have lots of systems running other HSM activity during the time
you are running Autobackup (like RECYCLE, EXPIREBV, etc.), then using
VSAM RLS for the HSM CDS's will also speed things up.
-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On
Behalf Of Richard Marchant
Sent: Friday, March 16, 2007 3:58 AM
To: [email protected]
Subject: HSM's Autobackup performance
Is there any parameters that can be tweaked to increase the performance
of HSM's autobackup?
A GTF trace shows that HSM backs-up datasets in 16K blocks.
Surely if HSM is using DFDSS as the datamover (default now) it should be
securing these datasets in 64K blocks at least.
Is it necessary to change any of the parameters in the DEVSUPxx member?
EG the COMPACT and TAPEBLKSZLIM
The HSM manuals have no information regarding the above.
Richard
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
