Comments on: Server Monitoring – Few Winners

By: Client-side charting for Leemba | Another guy named Mike

Client-side charting for Leemba | Another guy named Mike — Sat, 19 Dec 2009 21:49:16 +0000

[...] nobody’s heard from me in quite a while because I’ve been hard at work building my server monitor. Sorry, I suck. I have been, amongst other things, trying to get charting [...]

By: Ian McGowan

Ian McGowan — Fri, 15 May 2009 04:16:33 +0000

This is one of those rare cases where Sturgeon’s law fails – instead of 90% of everything being crap, 100% of monitoring solutions are…

By: Kevin

Kevin — Tue, 12 May 2009 03:14:28 +0000

Yeah, would never use java to monitor anything… Why have a monitoring tool that takes more resources that most of the services you would monitor?

Anyway, BB is painful. Very painful at times. The only “great” thing about it is that you can write just about any test you want and send alerts using the ‘bb’ binary.

By: Mike Johnson

Mike Johnson — Mon, 11 May 2009 19:41:42 +0000

Interesting, I must have missed that one. Looks like it was just released in October?

I’ve downloaded and installed. Pretty easy so far. Although it looks and acts a lot like Hyperic. I wrote about 5 or so of the list items thinking about Hyperic. :-) Jopr does have an improved interface.

So far seems to have some weird Firefox bug on on of my machines (can’t view the right pane). Not a big deal. Otherwise I like that it doesn’t prompt me to buy the enterprise version just to get user roles.

Unfortunately, adding a simple http check is throwing NullPointerExceptions in the log…

I like that the alerts configuration has an availability option — I never really grokked hyperic’s alert. It’s availability metric was a percentage so I ended up making alerts when it was less than 95% or something.

I’ll keep playing with it. Seems better than most but there are a few wish list items left.

By: Dino

Dino — Mon, 11 May 2009 16:20:11 +0000

You can try JOPR .. Open source java project .. from Redhat community

By: Shahbaz Javeed

Shahbaz Javeed — Mon, 11 May 2009 11:33:38 +0000

Looks like what you need is the opensource Community edition of Groundwork Monitor at http://www.groundworkopensource.com/community/downloads/. It uses nagios as the core scheduling engine and can utilize most (if not all) nagios plugins without any changes. As a satisfied user I can tell you that it functions well for what you need.

As far as the need to have an agent-less setup you can use SNMP (use the net-snmp package for RHEL/CentOS) as the agent to provide a lot of the information that you might otherwise need to login to the machine to get. I realize agent-less and “SNMP agent” don’t seem to jibe :) but I consider SNMP to be more part of the OS than anything else since you’ll find it implemented as part of Linux, Solaris, Windows, HP/UX and any other respectable server OS. Alternatively you can always stick to logging in using ssh to gather data but that won’t work on Windows machines.