-
Hi Marist
why would u look for smaller time slices in the statspack report.The statspack report is taken in a interval of 15 minutes..
regards
Hrishy
-
I am guessing that waiting on a logswitch is happening during that 15 minute window with not a lot of other activity going on. Therefore, the time spend waiting for a log switch may be high in comparison to other waits going on.
A smaller window will let you see what is happening between log switches. If you got still get high logswitch waits when no logswitch occurs, there is something else going on.
Jeff Hunter
-
Originally posted by hrishy
I dont have exact number but in the top 5 wait events .
BUT YOU DO! It's all there in the statspack, e.g.:
PHP Code:
Top 5 Wait Events
~~~~~~~~~~~~~~~~~ Wait % Total
Event Waits Time (cs) Wt Time
-------------------------------------------- ------------ ------------ -------
db file sequential read 9,747 7,168 45.34
control file parallel write 1,250 2,012 12.73
file open 441 1,379 8.72
enqueue 24 946 5.98
log file sync 557 715 4.52
-------------------------------------------------------------
Wait Events for DB: PROD Instance: prod Snaps: 992 -993
-> cs - centisecond - 100th of a second
-> ms - millisecond - 1000th of a second
-> ordered by wait time desc, waits desc (idle events last)
Avg
Total Wait wait Waits
Event Waits Timeouts Time (cs) (ms) /txn
---------------------------- ------------ ---------- ----------- ------ ------
db file sequential read 9,747 0 7,168 7 3.8
control file parallel write 1,250 0 2,012 16 0.5
file open 441 0 1,379 31 0.2
enqueue 24 0 946 394 0.0
log file sync 557 1 715 13 0.2
log file parallel write 644 0 538 8 0.2
control file sequential read 6,004 0 514 1 2.3
-
Originally posted by marist89
Therefore, the time spend waiting for a log switch may be high in comparison to other waits going on.
Jeff, I think this could be one of those cases where a ratio (58% of waits are due to X) is misleading. The real question is: how long are my users waiting because of X? Is it milliseconds, seconds, minutes, hours?
-
Originally posted by DaPi
Jeff, I think this could be one of those cases where a ratio (58% of waits are due to X) is misleading. The real question is: how long are my users waiting because of X? Is it milliseconds, seconds, minutes, hours?
True, could be.
Jeff Hunter
-
Hi Dapi
:-) Well i dont have acesses to raw data like that.Just graphs and pie charts :-)
regards
Hrishy
-
Originally posted by hrishy
Well i dont have acesses to raw data like that.Just graphs and pie charts
Aghhhhh! You've been got at by the marketeers! The "raw" data come from the standard "spreport".
P.S. how do you get the fancy stuff? I might want to impress my boss . . . .
-
Hi Dapi
There is a custom program written that will graph all the details to me with fancy HTML using applets etc..no acessses to raw data :-)
i guess we can easily write sucha prog using php and GD library :-)
regards
Hrishy
-
Hi Hrishy, Yet another GUI that stops you seeing the truth! If your "customisers" can't show you the TIME waited, it's time to send in Mr.Hanky (or Donald von Rumsfeld).
-
Originally posted by DaPi
Hi Hrishy, Yet another GUI that stops you seeing the truth! If your "customisers" can't show you the TIME waited, it's time to send in Mr.Hanky (or Donald von Rumsfeld).
Hi Dapi
hehe :-)
regards
Hrishy
Posting Permissions
- You may not post new threads
- You may not post replies
- You may not post attachments
- You may not edit your posts
-
Forum Rules
|
Click Here to Expand Forum to Full Width
|