Oracle instance often down

[rudhik]

Hi,

Could you please help ?

In my oracle server, there is 2 instances running. They use the same ORACLE_HOME, only different instance. As such,
both configuration similar.

Why is one of the instances often down while the other very stable ?

Thanks,
Rudhi

[shawish_sababa]

Hi

look at your alert log files.

please post the error message that you get, or your alert log files.

[rudhik]

Thanks for the reply.

Below is the error messages taken from alert log file :

---------------------------
Thread 1 cannot allocate new log, sequence 13659
Checkpoint not complete

(this error shows up very often)

----------------------------
Errors in file /vend/ow98+/oracle/rdbms/log/orow98_ora_5270.trc:
ORA-00449: background process 'LGWR' unexpectedly terminated with error 313
ORA-00313: open failed for members of log group of thread
ORA-00474: SMON process terminated with error
Wed Jun 27 01:15:39 2001
Errors in file /vend/ow98+/oracle/rdbms/log/orow98_ora_5270.trc:
ORA-00603: ORACLE server session terminated by fatal error
ORA-00449: background process 'LGWR' unexpectedly terminated with error 313
ORA-00313: open failed for members of log group of thread
ORA-00474: SMON process terminated with error
Wed Jun 27 07:28:49 2001
Shutting down instance (abort)

[shawish_sababa]

Hi

the most significant error reported here is the ORA-313 error.
this error indicates a failure to open a log file. this is not a case of i/o problem for in that case you would run into a different set of errors. becous, as you have writen, you have already used your db, i think that: a. your redo logs are switching too often, b. no idea at the moment...
so, increase the size of your redo logs and drop the old ones. as far as the info you posted here, this is the only idea i can come up with.

i fyou need more help, post the db version, os, if your db is based on upgrade, write about it. also, try to find a lgwr log and look for errors.

[hbk_emp]

make sure you v_$logfile or alertlog and check the sequenc no of the log so if they are > then 65000 then run
alter database open resetlogs commands and you lgwr wont fail any more
how much mem do you have on your box?

[sambavan]

What is the SGA Size on both the instances and what is your physical memory size on your system? What is your db version? What is your OS ?

[rudhik]

Thanks all.

FYI, this problem happened to my friend who is working at remote area. She asked me for help. Every solutions you posted here, needs time to try and response as I have to contact her first.

I don't have the info of SGA size at the moment.

The other info : 3GB RAM, Solaris 2.6, Oracle 7.3.4