Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Mailing Lists -> Oracle-L -> Enqueue wait time discrepancies in statspack
Can the list think of a good reason why statspack would show top event as enqueue (26,380 secs in a 15.4 minute period) while at the same time showing the top 10 enqueue activities adding up to 20 secs only?
Possibly of relevance
1) I'm wondering if the missing time could be related to AQ's 'skip locked' behaviour somehow. We have a couple of heavily used queues, one for requests and one for replies; the reply queue is polled by the client...
2) The only possibly relevant DB change I can think of prior to this happening is the addition of a couple of (as yet unused) tables with unindexed FK. These tables aren't used by the test (but the tables they reference are inserted and updated heavily - not deleted). Sadly I'm not allowed to add indexes to exclude this at the moment.
Thanks and regards Nigel
(Three node 9.2.0.5 RAC on Linux)
Top 10 Timed Events
~~~~~~~~~~~~~~~~~~ % Total Event Waits Time (s) Ela Time -------------------------------------------- ------------ ----------- -------- enqueue 95,747 26,380 75.60 buffer busy global cache 88,063 2,852 8.17 global cache null to x 74,094 1,075 3.08 CPU time 976 2.80 latch free 154,687 952 2.73 db file sequential read 164,339 914 2.62 buffer busy waits 74,369 860 2.46 global cache cr request 381,071 378 1.08 buffer busy global CR 131,858 201 .58 global cache open x 38,603 89 .25 Enqueue activity: Eq Requests Succ Gets Failed Gets Waits Time (ms) Time (s) -- ------------ ------------ ----------- ----------- ------------- ------------ TX 128,230 27,220 100,985 21,083 .93 20 US 7 7 0 7 13.29 0 FB 54 54 0 38 1.08 0 HW 122 122 0 18 1.17 0 CF 1,388 1,388 0 16 1.00 0 SQ 129 129 0 4 2.50 0 TM 247,657 247,655 0 9 .89 0 DR 1 1 0 1 3.00 0 TA 3 3 0 3 .33 0
-- http://www.freelists.org/webpage/oracle-lReceived on Tue Jun 13 2006 - 04:09:32 CDT
![]() |
![]() |