ora-610(2)

continue focus on ORA-610 errors issue, I got some clue from MOS, following as:

Cause

This is usually a resource problem. i.e. lack of resources at the OS level.

//beside, here are some workarounds you can take a look,but if the resource on os is going to exhausting, they are out of work.so , the key points is to enlarge the resource on OS.

Solution

You can try the following as solutions :

1. Increase the paramater JOB_QUEUE_PROCESSES.

2. Make sure Swap is correctly configured and increase it if necessary.

4 ~ 8 GBs of RAM, set Swap to at least double RAM size.

8~16 GBs of RAM, set Swap to double RAM.

16~32 GBs of RAM or more, set Swap at a rate of 0.75xRAM.

3. Make sure all the Kernel parameters are configured correctly and above minimum requirements, especially the maximum number of processes per non-root user.

4. Use the HCVE to check that all the minimum requirements are met.

Note 250262.1 – RDA 4 – Health Check / Validation Engine Guide

5. Check the ulimits and preferably set all to Unlimited.

6. Check the system log at the same time of the messages for errors/any indication of resource problem.

7. Check your SGA and PGA settings, make sure they are not set too low or too high. Preferably

PGA_AGGREGATE_TARGET is set at a rate of 2 MBs per process defined by the PROCESSES

parameter. An over allocation of resources could also cause the same.
One of the above solutions will most likely fix your problem but if none does, please log an SR with Oracle Support to assist you.

Hints:

You can monitor the PROCESSES usage at the database level using the V$RESOURCE_LIMIT view.

Also, if you monitor the database and see a lot of INACTIVE sessions, then you need to resolve that first before addressing the suggestions above, these sessions are lying idle eating up resources, and causing limitations, so the users need to make sure to exit their sessions gracefully, these sessions should be cleaned up regularly and perhaps consider using DCD and KeepAlive (Windows only).

Note 226202.1 – How to Set Keepalive on Windows NT/2000
Note 151972.1 – Dead Connection Detection (DCD) Explained

ORA-00610: Internal error code

一个客户的数据库遭遇报错,暂时到没有影响到客户端程序,发来了日志信息,让帮忙查看。
Tue Aug 14 14:22:20 2012
Errors in file c:oracleproduct10.2.0adminaaabdumpaaa_ora_3592.trc:
ORA-00610: Internal error code
Tue Aug 14 14:22:25 2012
Process J000 died, see its trace file
Tue Aug 14 14:22:25 2012
kkjcre1p: unable to spawn jobq slave process 
Tue Aug 14 14:22:25 2012
Errors in file c:oracleproduct10.2.0adminaaabdumpaaa_cjq0_3976.trc:
ORA-00600的internal报错经常见,610倒是第一次见到。在mos上大概搜了一下,发现与OS当时的资源消耗相关度比较大,问了客户当时的OS负载、操作等,当时os正经历IO负载较大的操作(大数据量的文件拷贝),资源相对紧张,但io的消耗增大也会对oracle产生影响吗?比较疑惑。
MOS上倒是有篇doc与症状很类似,如下:
ORA-610 INTERNAL ERROR CODE, KKJCRE1P: UNABLE TO SPAWN JOBQ [ID 436509.1]

Applies to:

Oracle Server – Enterprise Edition – Version: 11.1.0.6 and later   [Release: 11.1 and later ]
Information in this document applies to any platform.

Symptoms

When running a job from the scheduler on a very heavily loaded database you could receive an ora-610 error.  The job slave process creation takes a long time and raises a reported error as a fatal error in the background process.

The resulting trace file will have this information listed in it:

New process timed out
error 610 in job queue process
ORA-00610: Internal error code
Died during process startup with error 447 (seq=109)
OPIRIP: Uncaught error 447. Error stack:
ORA-00447: fatal error in background process

Cause

This is caused by lack of resources on the server.  If the system is extremely loaded, the timeout could be hit and the spawning error will occur. You will see this problem where a job slave process is not started is if the system is running out of resources (e.g. no memory left, max number of processes hit, etc.)

 

Solution

Ensure that there are enough resources on the server to satisfy all the scheduler/job queue requests.

症状非常类似,只是apply to :11.1.0.6让人疑惑,客户的环境是oracle10201
–EOF–