Bugzilla – Bug 7030
nimbusctl should wait longer to check process
Last modified: 2010-06-08 21:35:01
You need to
before you can comment on or make changes to this bug.
Five seconds is too short of a time to wait for simple initialization errors
that cause the process to die. I've experienced simple conf errors that show
up as OK for five seconds. There is a lot of code loading etc. before the
Nimbus code even starts to initialize and run all of its validation code.
Suggest the postStartDelay in nimbusctl.py be at least 15.
Made default 10 seconds, added a nimbus configure option "services.wait" so
that an admin/developer can change easily. The nimbusctl program will default
to 10 seconds if that configuration is not present (attempting at least to
guard against someone using an old nimbus-setup.conf file).
We agree that a better solution should be in place (some kind of
mechanism-based information from the services container that all is OK) but
this is better than the current 5 seconds at least.
Merged to master for Nimbus 2.5