z07a record count not going down
- Article Type: General
- Product: Aleph
- Product Version: 20
- Relevant for Installation Type: Dedicated-Direct; Direct; Local; Total Care
Description
xxx01 z07a record count is not going down. The count has been like this for several minutes:
xxx01@ALEPH20> select substr(z07a_rec_key,1,1), count(*) from z07a group by substr(z07a_rec_key,1,1);
SUB COUNT(*)
--- ----------
1 11377
2 1000
Resolution
As described in Article "What is the Z07A record? z07_level and z07a_level", the z07a is processed by both the ue_01_word_parallel and ue_01_z0102 processes. The z07a will not be removed until processing by both has occurred.
The count of z07a records "jumps" in increments of 1,000. Word-processing can take as long as 5 seconds per record, so it could take more than an hour for 1,000 records to be processed and for you to see a change in the count.
When ue_01_word_parallel has successfully processed a z07a record, it changes the z07a_level to '2'. The following SQL shows the current status:
SQL> select substr(z07a_rec_key,1,1), count(*) from z07a group by substr(z07a_rec_key,1,1);
After all the z07a's have been created (by ue_01_a), you should be seeing the count of 1's decreasing in increments of 1,000 every hour or two. (The count of 2's will remain at 1000 until all the z07a records are processed.)
If after four hours, there is no change in the count, then consult Article ("ue_01_z0102_index failing ").
If these don't help, then use util e/2 to stop the xxx01 ue_01 process. If that doesn't stop the ue_01_z0102 process (or ue_01_word_parallel process), then use the Unix "kill" command to stop it. Then do util e/1 to restart ue_01.
If this doesn't help, then try stopping/restarting Aleph and Oracle.
Additional Information
What is the Z07A record? z07_level and z07a_level
Category: Background processing
- Article last edited: 15-Aug-2016