PX Deq waitMan alive, I've been wrestling with the same thing the last few
days. The clients process always runs in less than two hours and now it's 12
hours. statspack should waits on PX Deq Credit: send blkd and CPU time. cpu
usage was at 99% but i/o waits were nill. Anyways, I've deduced that oracle
moved on to a different plan. Because I was testing out the query and it was
grabbing an irrelevant index. I flush the shared pool before their next
batch run and things were fine. It's all batch loads with adhoc queries so
the flush wasn't that big a deal..
- Original Message -----
From: David Sharples
To: oracle-l_at_freelists.org
Sent: Wednesday, June 09, 2004 8:46 AM
Subject: PX Deq wait
Hi, had a little problem today.
Had a process which had been running for over 24 hours (should take about 15
secs) it was waiting on PX Deq Credit: send blkd.
It was a job which did some inserts, statspack was telling me no i/o was
being done on the objects at all. Is it possible that the parallel slaves
had died so my process never knew it had finished, no trace files were
generated and nothing in the alert log, killing the process and restarting
it fixed it
Is there somewhere else I can look next time to see what has really
happened?
Thanks
Please see the official ORACLE-L FAQ:
http://www.orafaq.com
To unsubscribe send email to: oracle-l-request_at_freelists.org
put 'unsubscribe' in the subject line.
--
Archives are at
http://www.freelists.org/archives/oracle-l/
FAQ is at
http://www.freelists.org/help/fom-serve/cache/1.html
Received on Wed Jun 09 2004 - 08:03:49 CDT
