On 29/11/16 08:24 +0100, Ulrich Windl wrote:
> Some servers first fork and exit, virtually being successful
> immediately, while the child (the "server loop") could die a moment
> later. Finding the perfect time to wait for the server dying is kind
> of black magic ;-)

That's exactly why I thing that's one of the aspects systemd moved
service startup design a step forward with its "notify" type of service
-- only the daemon author is fully capable/educated to narrow down
the exact point in the daemon's life cycle where it's supposed to
be "fully up, ready to service", marking that place with sd_notify(3)
call.  Everything else is just a heuristic, sometimes unreliable.

Jan (Poki)

Attachment: pgpBSDOK0BrM1.pgp
Description: PGP signature

Users mailing list: Users@clusterlabs.org

Project Home: http://www.clusterlabs.org
Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
Bugs: http://bugs.clusterlabs.org

Reply via email to