Thanks to all who helped me with this one, looks like we have * bug *. OWS
is working on resolution and probably a patch. Once I get a clear research
note from dev, I'll update the list if anyone is interested.
Raj
__
Rajendra Jamadagni
Check MetaLink Doc ID: Note: 176129.1 ALERT: "LATCH FREE" And
"FREE_BUFFER_WAITS" Cause Performance Degradation/Hang
Problem affects 9.0.1.2 and sounds like what you're experiencing.
HTH,
david
David B. Wagoner
Database Administrator
Arsenal Digital Solutions Worldwide, Inc.
8000 Regency Park
This query was totally customized for our production database.
We run this query every 2 secs
This is querying the main knowN events which causes sometimes the database
to freeze
especially the latch free wait you can customize it as you want.
the format comes out like below
SIDEVENT_WITH_
AIL PROTECTED]]
> Sent: Tuesday, March 12, 2002 10:14 AM
> To: Multiple recipients of list ORACLE-L
> Subject: RE: DB freezes ... no indications of any kind
>
>
> If you are able to login and query the session_Wait
> try using the below said query
>
> select s.IND
It's a beautiful query which only misses "FROM" clause. I'll save it
for later use. Much later.
> -Original Message-
> From: Arun Chakrapani [mailto:[EMAIL PROTECTED]]
> Sent: Tuesday, March 12, 2002 10:14 AM
> To: Multiple recipients of list ORACLE-L
And how would I interpret the output of this query?
Thanks Arun
Raj
__
Rajendra Jamadagni MIS, ESPN Inc.
Rajendra dot Jamadagni at ESPN dot com
Any opinion expressed here is personal and doesn't reflect that of ESPN Inc.
QOTD: Any
If you are able to login and query the session_Wait
try using the below said query
select s.INDX
SID,u.KSUUDNAM||','||u.KSUSEUNM||','||u.KSUSEMNM||'('||decode(KSUSSOPC,2,'LA
TCH FREE',12,'ENQUEUE',94,'db file sequential read',
95,'db file scattered read',144,'LIBRARY CACHE PIN',75,'buffer busy wa
Thanks john,
The dbms_job has problems on 9012, so I am switching to good old cron ...
basically I don't have the statpack info at all ... but I am starting it at
10am today.
Raj
__
Rajendra Jamadagni MIS, ESPN Inc.
Rajendra dot Ja
Thanks Stephan,
No it is not a problem with archiving at all, the filesystem assigned for
archiving has may more GB of space than it needs.
Raj
__
Rajendra Jamadagni MIS, ESPN Inc.
Rajendra dot Jamadagni at ESPN dot com
Any opinion
When your database freezes can you log in as internal thro svrmgrl and see
what is happening on v$session_Wait
-Original Message-
Sent: Tuesday, March 12, 2002 9:19 AM
To: Multiple recipients of list ORACLE-L
Thanks john,
The dbms_job has problems on 9012, so I am switching to good ol
Thanks Waleed,
Actually this happened again last night ... it took me 2 seconds to log in,
so no sqlnet problem. The system was dumping trace files like crazy, yet
Spotlight didn't show any trouble spots. Let me see what Oracle has to say
on this the TAR is already in progress.
Raj
Thanks Arun
Raj
__
Rajendra Jamadagni MIS, ESPN Inc.
Rajendra dot Jamadagni at ESPN dot com
Any opinion expressed here is personal and doesn't reflect that of ESPN Inc.
QOTD: Any clod can have facts, but having an opinion is an art
Raj,
> Well, I'd say about 3 log switches during 15 minute period
> before and after
> the freeze, nothing crazy ... that's what is confusing ...
Could you post the Statspack report for the period where you experienced
this 'hang' as well as one during 'normal' periods? If you don't want to,
ha
like put a cron job which keeps connecting to the database if there a probs
by any chance a freeze see to that in the script it runs ORADEBUG COMMAND
WHICH WILL TAKE A DATABASE SYSTEM DUMP
I have a feeling u know the command for ora debug
If not please use the following script
I use the below sai
You have to find out:
If it's only a problem for users/programs that use sqlnet+listener to
connect to the server.
if you can connect locally to Oracle while others can not connect remotely
then could be network traffic problem.
If you can connect internal locally but can not connect locally throu
Can you please elaborate on that?
Raj
__
Rajendra Jamadagni MIS, ESPN Inc.
Rajendra dot Jamadagni at ESPN dot com
Any opinion expressed here is personal and doesn't reflect that of ESPN Inc.
QOTD: Any clod can have facts, but havin
-4350
Mobile: (248) 408-2918
Email: [EMAIL PROTECTED]
Web:www.compuware.com
-Original Message-
Sent: Monday, March 11, 2002 3:58 PM
To: Multiple recipients of list ORACLE-L
Subject:RE: DB freezes ... no indications of any kind
<< File: ESPN_Disclaimer.txt >> T
Why dont you sent an event which monitors the database and takes a system
state dump during the freeze thro ORADEBUG
>From that I hope u can find out where it is bombing
-Original Message-
Sent: Monday, March 11, 2002 3:58 PM
To: Multiple recipients of list ORACLE-L
Thanks Waleed,
Well
Thanks Waleed,
Well, I'd say about 3 log switches during 15 minute period before and after
the freeze, nothing crazy ... that's what is confusing ...
Raj
__
Rajendra Jamadagni MIS, ESPN Inc.
Rajendra dot Jamadagni at ESPN dot com
A
Check the frequency of Log Switches, specially that time the db froze
Waleed
-Original Message-
Sent: Monday, March 11, 2002 8:43 AM
To: Multiple recipients of list ORACLE-L
Hi all,
this is something new on our 9012 db. Once in two days, the DB will freeze
up for up to 2-3 minutes ...
Raj - If you run STATSPACK with the statsauto script, then it runs every 15
minutes, which would be too large an interval for your purposes. However, if
you
execute statspack.snap;
then it executes immediately. When the problem occurs, execute it. If you
are concerned, keep a window open
Dennis,
the problem is, stats pack s schedules to run every 15 minutes, the problem
is I want to know what happens during these 2-3 minutes why oracle is
not telling me anything ...
Thanks
Raj
__
Rajendra Jamadagni MIS, ESPN I
Raj - Consider running statspack. If you have it configured, you can just
enter:
connect perfstat
executes statspack.snap;
This should help you to see what is going on inside Oracle.
Dennis Williams
DBA
Lifetouch, Inc.
[EMAIL PROTECTED]
-Original Message-
Sent: Monday,
23 matches
Mail list logo