Oracle on RHEL4 with ASM problem

[Jack Challen]

Lunt, Nick wrote:
> h1:35:respawn:/etc/init.d/init.cssd run >/dev/null 2>&1 </dev/null
> 
> When ASM throws this error I have to comment out this entry in inittab, run 'init q', uncomment the inittab entry and run 'init q' again.
> This fixes the problem.

I don't have a copy of this script kicking around here, but it sounds 
like it's timing out or locking somehow. Seems like they (Oracle) are 
using a cross between SysV init scripts and init(1) handling; have you 
tried running either:

/etc/init.d/init.cssd status   #(/may/ not work)
/etc/init.d/init.cssd restart  #(should work)

as root?

Failing that, you may able to hack that script to put e.g. "sleep 5" at 
the top -- this is a dirty hack; hopefully running the script by hand 
should give you some more information.
If the sleep trick fixes it, then I'd suggest that that script is 
restarting too quickly for init(1) to assume that it's successful - you 
should see messages in /var/log/messages that say something like
init: process h1 respawning too quickly -- disabled for 5 minutes.


hth
jack


-- 
Jack Challen
Technical Consultant, OCSL.
http://www.ocsl.co.uk/