Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> How can I troubleshoot without informations ?

How can I troubleshoot without informations ?

From: Dusan Bolek <pagesflames_at_usa.net>
Date: 11 Dec 2001 08:21:59 -0800
Message-ID: <1e8276d6.0112110821.5d46128e@posting.google.com>


We have problem on one of our servers. Everything is more complicated because this server is NT box and I can get in, because of security policies.
So I've tried to do troubleshooting with information from second (maybe third) hand. I love this situation. Clients have got well-known ORA-12500 error. This Oracle 8.0.5.1.1 server need to handle 800 concurrent connections in dedicated mode, which is something not very good for NT platform. The problem is probably in insufficient memory. However, I cannot tell this for sure, because some clever head decided to switch off logging of listener actions. So there is no listener.log, which is one of my favourite Oracle log files. :-(
Server has 2GB of physical RAM, but it seems that Oracle is using only 1GB of it. That's strange because there is a lot of free memory available. The only way how they can achieve eliminating ORA-12500s is decrease of SGA which left more memory available for PGA. However performance trade-offs are very serious (nothing suprising). My situation is pretty bad, because I can get my hands on that very server, I'm not very confident about capabilities of DBA taking care about it and I do not know NT environment very well, especially in memory allocation limits.
I know that there should be 2GB limit for user processes in NT, but 1GB seems for me too low value to has problems. If someone can help me, it will be appreciated.

Some interesting parameters from init.ora: control_files = (E:\OraCtrl\DB10\ctrl_DB10_01.odf,E:\OraCtrl\DB10\ctrl_DB10_02.odf,E:\OraCtrl\DB10\ctrl_DB10_03.odf) <- very good idea having
all control files together. :-(

db_block_buffers = 60000
db_block_size = 4096
db_files = 254
dml_locks = 500

log_buffer = 163840
log_checkpoint_interval = 25000
open_cursors = 200
processes = 1600 <- was 2000
shared_pool_size = 25000000 <- was 30000000 then 15000000 sort_area_size = 258048
MTS disabled, all running in dedicated mode

P.S. In alert log is nothing interesting. However I'm not sure if I have the log from that time. :-(

--
_________________________________________

Dusan Bolek, Ing.
Oracle team leader

Note: pagesflames_at_usa.net has been cancelled due to changes (maybe we
can call it an overture to bankruptcy) on that server. I'm still using
this email to prevent SPAM. Maybe one day I will change it and have a
proper mail even for news, but right now I can be reached by this
email.
Received on Tue Dec 11 2001 - 10:21:59 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US