Short version: If you are z/OS JES2 1.7 or above, don't drain spool volumes. APARs OA17249 and OA20195 (new, see "additional related symptoms" within the APAR text) describe problems which can occur if you $P a spool volume which is not known to be empty. The below workaround is suggested.
Brian On Mon, 5 Mar 2007 09:19:19 -0600, Brian Peterson wrote: >There are problems in z/OS 1.7+ JES2 where SPOOL volumes can drain with >valid occupied track groups still remaining. > >One of these problems is described by OA17249. The fix for OA17249 only >corrects jobs submitted AFTER the fix is applied. > >Another problem was just discovered (no APAR number yet). > >Both of these problems deal with issues related to something called >the "SPOOL used mask". This is an attribute of each job, and is incorrect >due to reasons described in OA17249 and in the new APAR about to be >created. Because the mask is wrong, JES2 "thinks" spool volumes are empty >when in fact they still contain parts of valid jobs. > >The bottom line is - do not $P spool volumes in JES2 1.7 or 1.8. Instead, >follow the below procedure: > >Step 1) $T SPOOL(volser),SYSAFF=-ANY > This will perform the same function of $P SPOOL(volser), but without > draining the spool volume. > >Step 2) Take note of the exact date / time of step 1. > >Step 3) Wait until every job with a RD-DATE / RD-TIME (from SDSF ST display) > which was submitted before Step 2's date / time has purged. On my > system, this took several weeks. > >Step 4) Once it is known that no job exists on SPOOL which pre-dates the > date / time of Step 2, it is now safe to $P SPOOL(volser). > >Brian > ---------------------------------------------------------------------- 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
