recompile invalid objects crashes database

[bookouri]

I've never had anything like this happen before. Periodically I run some scripts to identify and recompile invalid objects. When i did my recompile invalid objects today, it crashed the database and i had to perform a "recover database" to get things back up and going. I always understood that there was no danger in recompiling invalid objects. I havnt started researching the error written in the log file:
PMON: terminating instance due to error 476
Mon Apr 02 10:26:54 2007
Instance terminated by PMON, pid = 3304
Have i been wrong all this time about the lack of danger in recompiling invalid objects?

any comments would be appreciated.

 

[hoinz]

What makes you believe that the crash was caused by recompiling?
Could be just a coincidence, I presume.
Or could be an old problem with Oracle datafiles, corrupt blocks or the like.
Any hints in Oracle's alert log, or in operating systems logs?

If you like playing risky games, you might try recompiling once again. But I would run a backup before doing more trials.

 

[bookouri]

i thought about it being a coincidence, but it really is hard to believe it coinciding THAT close to when i ran my script. No sign of anything else in the event logs or anywhere. Im hoping its just a one time thing. What I really want to make sure of is that under normal circumstances recompiling invalid objects shouldnt be able to crash the database.

 

[SantaMufasa]

Book,

The only recompilation that I can imagine that might crash the database would be a re-compile of an instance-dependent object or module. Does your recompilation script attempt to recompile objects that SYS owns?

Mufasa
(aka Dave of Sandy, Utah, USA)
[I provide low-cost, remote Database Administration services: www.dasages.com]

 

[bookouri]

Ill check to make sure, but Im pretty sure the scripts select only the owner's objects. I know for sure they were not run under sys's logon.