Make upstart script respawn consul on crash

[tgwizard]

This could be configurable (whether to respawn, and the count/interval/unlimited), but I think it could be a future PR.

Is there a reason to not respawn if consul crashes?

[coveralls]

Coverage remained the same when pulling de2965f on tgwizard:upstart-respawn into 7ea99ed on johnbellone:master.

[lyrixx]

👍 for the features

[tgwizard]

@johnbellone what do you think?

[johnbellone]

I personally don't see anything wrong with this. Since its a change in behavior we should just bump the minor revision.

Any thoughts @reset ?

[johnbellone]

@tgwizard Is there any logic in Consul that is destructive on a restart? Assuming it crashed we want to make sure that we're not wiping out data. Otherwise, it 👍 !

[tgwizard]

@johnbellone there seems to be no issue for me. I can sudo kill -9 $PID, the consul process dies, and then restarts and rejoins the cluster. I have rejoin_after_leave set to true in my consul config though, which is (I think) what causes the rejoin to work. Or perhaps not, since no leave was triggered.

[lock]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.