From: Martin K. <kn...@kn...> - 2008-02-29 08:00:45
|
----- Original Message ---- > From: john allspaw <jal...@ya...> > To: Martin Knoblauch <kn...@kn...>; Brad Nicholes <bni...@no...>; Peter Mui <pm...@gr...>; gan...@li... > Sent: Friday, February 29, 2008 1:39:17 AM > Subject: Re: [Ganglia-developers] Ganglia 3.1 wish list... > > My sincerest apologies for not making it over today, I will be showing my mug > there tomorrow. > > My vote on this: > > " - Add some event notification mechanism if metrics go over a limit. But do we > want to implement another Nagios?" > > No, no, no, no and no. :) > > All opinion here, but: > > I think to add event notification would be a major mistake, and would pull > attention off of what makes ganglia awesome, > which is the non-judgemental recording of system metrics. There already exist a > lot of ways to get ganglia's metrics into Nagios, > which has all of the bits that you'd want for a notification system. > > There are so many more cool/good/appropriate things to get into ganglia than > event notification. > -- john allspaw > flickr.com > Hi John, I am actually pretty much of the same opinion. Thats why I put the "But, ..." into my proposal. I have seen the request before and always thought "do we really need another Nagios". Similar with gexec/authd. Do they really belong into the monitoring-core? How does "integrating" them help recording the metrics? Cheers Martin > |