ctdb event scripts

[Martin Schwenke]

On Mon, 14 Nov 2016 15:53:40 +1100, Amitay Isaacs <amitay at gmail.com>
wrote:

> On Mon, Nov 14, 2016 at 1:17 PM, Martin Schwenke <martin at meltin.net> wrote:
> 
> > On Mon, 14 Nov 2016 12:42:28 +1100, Amitay Isaacs <amitay at gmail.com>
> > wrote:
> >  
> > > On Sat, Nov 12, 2016 at 9:07 PM, Martin Schwenke <martin at meltin.net>  
> > wrote:

> > > > > > Cool feature idea: implement an option to scriptstatus that
> > > > > > shows the output of the last failure.  This would require
> > > > > > daemon support,  since that's where the information comes
> > > > > > from.  

> > > > > Yes - that is a cool idea.  

> > > > Let's see if Amitay agrees and wants to add it to his new event
> > > > daemon.  :-)  

> > > What's so cool about it?  It's in the logs already when the node
> > > becomes UNHEALTHY.  I don't think there is any point adding it to
> > > eventd.  

> > So you don't need to grovel through the logs for the simple cases.

> > > If we want to provide a hook to capture that information, then it
> > > can be done via notify script.  

> > You could do it for any event.  It doesn't have to be tied to
> > something that generates a notification.

> > If you've had "startup" problem and you think they're now solved
> > then you can check with "ctdb scriptstatus startup lastfail" or
> > similar. Alternatively, if some higher-level monitoring software
> > has logged an unexpected fail-over then you could use "ctdb
> > scriptstatus monitor lastfail" for an initial look.  It might tell
> > you enough...

> What about timeout? Is that always a failure from eventd's point of
> view?
> 
> We have weird rules for handling timeout as failure in CTDB.  Not all
> timeouts are considered failures.

Yeah, good question.  As discussed, let's leave the timeout
handling/counting outside eventd.  If a script times out then that is a
failure from eventd's point-of-view.

Bring on lastfail and lastpass!  :-)

peace & happiness,
martin