RE: uptime bug

[rory]

>> It seems that the "uptime bug" is not a bug at all. On my systems that
>> I'm running on, they get their time set from satellite, and some of
>> the monitored processes come up before the satellite connection comes
>> up and can set the time, so they get a bogus time.
>
> :-) I wouldn't be surprised if you have GPS for your machines as well,
> so you know where the are at all time.
>
>> When should we release 3.1?
>
> Well, I'm not sure that Christian is completely ready with the SSL
> stuff and how is the status on the dependency functionality? I've been
> quite busy so I haven't had time to test out any of the new stuff. But
> lets say that we aim for a beat release at the end of next week, does
> this sounds reasonable?
>

As far as I know, the dependency stuff is working as designed. I have a
fairly complex test setup, and it handles it just fine.
I have a dependency chain that looks like:

database
resource manager
database abstraction
resource sheduler
actual program

In my case, if the resource manager is stopped, the dependenat programs
usually crash (or worse) and monit seems to be doing the right thing for
me. (as of my last checkin, that is)