IBM Mainframe Forum

Hi,
i hope this is posted in the right section - i am a new mainframe storage administrator, having been promoted from mainframe senior operator so i am still learning my trade !! So please bear with me -

i had an error last night and i am wondering if someone could help -

SP01) 21:52 DFHSM ARC0542I SMALL DATA SET PACKING DATA SET
(SP01) ARC0542I (CONT.) MSHSM.SMALLDS.VHSMD55 IS FULL

During the migration of a data set to small data set packing (SDSP) data set dsname, an attempt to write the migration copy to the SDSP data set failed due to a lack of space for the records

The 3 SDSP's we have are
MSHSM.SMALLDS.VHSMD54
MSHSM.SMALLDS.VHSMD54.DATA 3750 ? 1
MSHSM.SMALLDS.VHSMD54.INDEX 75 ? 1
MSHSM.SMALLDS.VHSMD55
MSHSM.SMALLDS.VHSMD55.DATA 3750 ? 1
MSHSM.SMALLDS.VHSMD55.INDEX 75 ? 1
MSHSM.SMALLDS.VHSMD56
MSHSM.SMALLDS.VHSMD56.DATA 18600 ? 104
MSHSM.SMALLDS.VHSMD56.INDEX 40 ? 6

i understand i need to reorg the dataset using idams command - but does anoyne have the command please?
Thanks in advance for any help given

If you are using RAID6 DASD then the need for SDSP is removed because of the difference of hardware compression within the DASD subsystem is quite different from RAID5 . I know when I first encountered RAID6, many many years ago, we did some experiments and found this to be the case.

However, if you want to stick with SDSP you need to use IDCAMS to REPRO the file to a backup, DELETE / DEFINE the cluster, and then REPRO from the backup to the cluster.

Another option is IDCAMS EXPORT, DELETE the cluster, IMPORT the cluster. This is my preferred method as you do not need to redefine the cluster as it is done automatically via the IMPORT process. If you want to change the space attributes of the SDSP then use the REPRO method above.

As I no longer bother with SDSP, and have not done so in many many years, I can not recall if we had to stop HSM or not. Probably safest to stop it before you start

brilliant reply - thank you.

I've just actione dit and it worked fine - thanks again for speedy reply.

one last question - after the import has finished - will i have to issue an ADDVOL command to activate it??
i did not stop HSM to export/imort.

You would only need to do an ADDVOL if you removed the ML1 volume from HSM before you started.

The only reason I used to stop HSM was in case someone wanted to recall one of the datasets stored in the SDSP