Dave Miner wrote: > Darren Reed wrote: > ... > >> >> Personally, I don't care if the service is still doing >> its start method or something else. If disable (and -t) >> are the only proper hooks to stop a service then that >> is what I expect it to do, not futz around and pretend. >> > > Did you try "-s" on svcadm (clearly you didn't in the script, but > outside of it)? That should get you closer to what you're looking for > in terms of failure semantics.
Using "-s" just means it won't return until it completes which could be how long? To my thinking, "disable" (for a service like NTP) should take a trivial amount of time to execute, regardless of its state. Maybe what should be said here is that SMF isn't trying hard enough to disable a service. I didn't dig into the why it doesn't try harder. Is it waiting for it to become "online" (properly) so it can then "offline" it? I hope not. If it is processes like ntpdate which are the problem then they also need to be fixed to be more responsive to SMF IPC, but to my recollection, ntpdate will terminate pretty easily with a ^C. Darren
