• Article Type: General
• Product: Aleph
• Product Version: 20, 21, 22, 23

Description:
Yeterday morning the system was pegged with Oracle seeming to be in some loop. I finally pin the problem as p_cir_32 running in for ABC50. After killing a bunch of jobs and restarting the system I ran cir-32 for abc50 again. It is still running, almost 24 hours now. The output file is growing very rapidly.

I ran cir-32 manually on a couple of other, smaller colleges, and it also took a while and built a quite large output file. Looking at a couple of the records in that case, it almost looks like the job is attempting to re-do the scrub for every record in the history. At least, the couple I looked were transactions that had completed about 3 years ago

cir-32 runs every night on production with days set to 2 and most of the major transactions set to be scrubbed. Therefore, the database shouldn't have a lot of un-scrubbed records.

Resolution:
As suggested in article p_cir_32: too many records older than 999 days. Can we use SQL? (KB 8192-4263), it may be simplest to handle the z35, z36h, and z37h with SQL. (These tables do not display online.) You could run SQL to update *all* of the z35_id, z36h_id, and z37h_id to "SCR2010mmdd".

This SQL update should be a one-time thing. It should *not* be necessary on an ongoing basis.

The z309 is a bit more complicated since it does display in the GUI and you want/need to preserve the actual z309_id in some cases. As described in the article p_cir_32 deletes log records for non-returned loans and unpaid cash (KB 16384-18434), rather than running p_cir32 for the z309, we suggest running p_cir_78.

• Article last edited: 10/8/2013