Hi,

DB2 V7 on z/OS

1. I have a reorg tablespace job with shrlevel NONE that failed in sort with an E37 on the SORTOUT file. All files were coded with MOD,KEEP,KEEP and allocated and cataloged in a previous step.

To restart with a RESTART(PHASE), SORTOUT file needs to have more space. I suppose it is ok to delete the old SORTOUT file and create a new one with more space i.e. no of rows * no of indexes * (longest key +9)

2. And if the E37 happens on SYSUT1, I suppose the same process as above is valid here also.

Could somebody please confirm this restart process is ok? Thanks in advance!

consider using the REBUILD INDEX utility instead of restarting the REORG TABLESPACE if it fails in the SORT phase.

I initially thought so. But it is a fairly large table with a few million rows. Rebuild index would have unloaded the index keys again for sort instead of using the SYSUT1 file already present.

Besides, it is a scheduled job with many reorg steps, and is it a good idea to manually do a rebuild index and skip the step in production job?