Services timeout and startup order.
263960Apr 14 2008 — edited Apr 16 2008Hey all, after a reboot the only service thats starts is the VIP. The rest are online but in an unknown state.
In my crsd.log file I see alot of time outs, with the first one failing ora.<nodename>.ASM1.asm
I have enabled debugging via the service and in racgwrap via the parameter USRORA_DEBUG but my ASM log doesn't even get written to on reboot. In my crsd.log:
Attempting to start `ora.raclinux`.ASM`.asm` on member `raclinux`
Action Script /home/oracle/product/10.2.0/asm1/bin/racgwrap(start) timed out for ora.raclinux1.ASM1.asm! (timeout=2000) note: in increased the timeout from 600
StartResource error for ora.raclinux1.ASM1.asm error code = -2
Action Script /home/oracle/product/10.2.0/asm1/bin/racgwrap(stop) timed out for ora.raclinux1.ASM1.asm! (timeout=2000)
StopResource error for ora.raclinux1.ASM1.asm error code = -2
X_OP_StopResourceFailed : Stop resource failed (File: rti.cpp, line: 1698
`ora.raclinux1.ASM1.asm` has experienced an unrecoverable failure
As soon as I reboot I tail my crsd.log file and within seconds after the first lines appear above it timesout... No where close to 2000 seconds. Also, none of the asm log files get touched.
If I do a crs_stop all , then crs_start all I get dependency errors on ASM. However, if I shutdown everything (crs_stop -all) and start everything up with srvctl it starts fine. ie.
srvctl start nodeapps -n raclinux1
srvctl start asm -n raclinux1
srvctl start database -d RACDB
everything works. I believe its not starting up in the correct order... How do I modify the startup order in clusterware...?