Comments on: Gandalf! Monitor Your Machines with Hobbit http://www.linux-mag.com/id/6138/ Open Source, Open Standards Sat, 04 Jul 2009 18:22:12 +0000 http://wordpress.org/?v=2.0.11 by: Rob Enders http://www.linux-mag.com/id/6138/#comment-2772 Wed, 10 Jun 2009 13:07:11 +0000 http://www.linux-mag.com/id/6138/#comment-2772 I have used Nagios since 2001 and more recently, the Opsview version. Opsview has improved the SNMP integration of Nagios. I have compared Nagios with Zenoss, BB and Zabbix, I believe Nagios is more robust and scalable. Custom scripts are a cinch with Nagios plugins. I run Opsview within a vm, monitor aprox 200 services and 75 hosts so not a huge system.. <a href="www.opsview.org" rel="nofollow">Home of Opsview</a> I have used Nagios since 2001 and more recently, the Opsview version. Opsview has improved the SNMP integration of Nagios. I have compared Nagios with Zenoss, BB and Zabbix, I believe Nagios is more robust and scalable. Custom scripts are a cinch with Nagios plugins.
I run Opsview within a vm, monitor aprox 200 services and 75 hosts so not a huge system..

Home of Opsview

]]> by: daniels http://www.linux-mag.com/id/6138/#comment-1849 Thu, 11 Dec 2008 13:21:08 +0000 http://www.linux-mag.com/id/6138/#comment-1849 Im using <a href="http://www.pandorafms.org/" title="Pandora FMS" rel="nofollow"> and Im quite happy with it. It was easy to install/use/configure. It has a SNMP interface and agents for linux, windows, solaris, etc. You can even monitor without angent at all. Realy good. Im using and Im quite happy with it. It was easy to install/use/configure. It has a SNMP interface and agents for linux, windows, solaris, etc. You can even monitor without angent at all. Realy good.

]]> by: andrea ferraris http://www.linux-mag.com/id/6138/#comment-1844 Tue, 09 Dec 2008 09:37:23 +0000 http://www.linux-mag.com/id/6138/#comment-1844 What about Monit (http://mmonit.com/monit/) Fo me it worked great. What about Monit (http://mmonit.com/monit/)
Fo me it worked great.

]]> by: Marcel Sugano http://www.linux-mag.com/id/6138/#comment-1842 Mon, 08 Dec 2008 19:03:57 +0000 http://www.linux-mag.com/id/6138/#comment-1842 All your comments about all NMS being fittest than Nagios for your environment do not consider monitoring about 1000 servers. Any hosting provider employee here uses zabbix/zenoss/groundwork/hobbit/bb on thousand-fold servers data-center? I don't think so. When scaling is what matter, nothing scale more than Nagios + Cacti(with boost plugin) as for NMS+trends_graphing solution. All your comments about all NMS being fittest than Nagios for your environment do not consider monitoring about 1000 servers. Any hosting provider employee here uses zabbix/zenoss/groundwork/hobbit/bb on thousand-fold servers data-center? I don’t think so. When scaling is what matter, nothing scale more than Nagios + Cacti(with boost plugin) as for NMS+trends_graphing solution.

]]> by: ashusethi http://www.linux-mag.com/id/6138/#comment-1840 Mon, 08 Dec 2008 17:24:13 +0000 http://www.linux-mag.com/id/6138/#comment-1840 I find Nagios is a better one though, had extensively used it, specially there is alot of development of supporting applications going on.. as one i remeber was a desktop alert client for windows, it used to show a pop up on my windows host that something is wrong with some server. Zabbix may be another good bet with a filtered GUI. Cheers ash I find Nagios is a better one though, had extensively used it, specially there is alot of development of supporting applications going on.. as one i remeber was a desktop alert client for windows, it used to show a pop up on my windows host that something is wrong with some server.

Zabbix may be another good bet with a filtered GUI.

Cheers
ash

]]> by: jjabusch http://www.linux-mag.com/id/6138/#comment-1205 Fri, 20 Jun 2008 19:33:21 +0000 http://www.linux-mag.com/id/6138/#comment-1205 I don't know about Zabbix, but I can speak to Nagios and Zenoss. Nagios is okay for doing simple testing and monitoring / alerting. However, when establishing SNMP tests, it's more challenging. Zenoss, however, was a little easier to setup than Nagios, and its autodiscovery works just fine. Plus, for devices it doesn't already know (we use a lot of Nortel equipment) - I just had to acquire the Nortel MIBS, and import them in through a menu option. Nagios has some basic charting abilities plus the ability to dump stats to other tools, although you may have to get creative. Zenoss will chart just about any stat it captures for you, with little effort. With that all said, I run both. Nagios is primarily just an alerting tool, because the stats and the ability to delve deeper into equipment is built into Zenoss, and I'd have to do a lot of scripting to make it work in Nagios. Plus, Nagios' dashboard-like functionality is still a little ahead of Zenoss' dashboard, IMO. I don’t know about Zabbix, but I can speak to Nagios and Zenoss. Nagios is okay for doing simple testing and monitoring / alerting. However, when establishing SNMP tests, it’s more challenging.

Zenoss, however, was a little easier to setup than Nagios, and its autodiscovery works just fine. Plus, for devices it doesn’t already know (we use a lot of Nortel equipment) - I just had to acquire the Nortel MIBS, and import them in through a menu option.

Nagios has some basic charting abilities plus the ability to dump stats to other tools, although you may have to get creative. Zenoss will chart just about any stat it captures for you, with little effort.

With that all said, I run both. Nagios is primarily just an alerting tool, because the stats and the ability to delve deeper into equipment is built into Zenoss, and I’d have to do a lot of scripting to make it work in Nagios.

Plus, Nagios’ dashboard-like functionality is still a little ahead of Zenoss’ dashboard, IMO.

]]> by: codemunkee http://www.linux-mag.com/id/6138/#comment-1180 Thu, 12 Jun 2008 18:37:14 +0000 http://www.linux-mag.com/id/6138/#comment-1180 I'll stick with Nagios and NRPE. Splunk is also awesome if you're wanting to monitor syslogs. And for Gentoo Admins, like myself, it becomes more clear what to use when you run: emerge -s hobbit and emerge -s nagios I’ll stick with Nagios and NRPE. Splunk is also awesome if you’re wanting to monitor syslogs.

And for Gentoo Admins, like myself, it becomes more clear what to use when you run:

emerge -s hobbit

and

emerge -s nagios

]]> by: Corne Beerse http://www.linux-mag.com/id/6138/#comment-1167 Wed, 11 Jun 2008 08:25:00 +0000 http://www.linux-mag.com/id/6138/#comment-1167 The nice thing about all these monitoring tools (hobbit, bb, nagios, zabbix) is that they have an agent for _most_ systems. How about the other systems? Do they come with a snmp interface? Is this a self-discovery interface? does it accept mib-files? Because in the end, you like to monitor _all_ systems, even the ones for which there is no agent and the ones that cannot handle an agent, like printers, switches, routers and other attached stuff. Look at HP with their server-agents: they add entries to snmp which is already available in the os. Then your own snmp-based monitoring can also monitor the HP-special parameters. That's the way it should be: Use an available standard, then you can get much higher. And for the microsoft operating systems: wmi is not the standard. The nice thing about all these monitoring tools (hobbit, bb, nagios, zabbix) is that they have an agent for _most_ systems. How about the other systems? Do they come with a snmp interface? Is this a self-discovery interface? does it accept mib-files? Because in the end, you like to monitor _all_ systems, even the ones for which there is no agent and the ones that cannot handle an agent, like printers, switches, routers and other attached stuff.

Look at HP with their server-agents: they add entries to snmp which is already available in the os. Then your own snmp-based monitoring can also monitor the HP-special parameters. That’s the way it should be: Use an available standard, then you can get much higher.

And for the microsoft operating systems: wmi is not the standard.

]]> by: Brian Majeska http://www.linux-mag.com/id/6138/#comment-1160 Fri, 06 Jun 2008 07:28:17 +0000 http://www.linux-mag.com/id/6138/#comment-1160 Hobbit smokes Nagios. I used Nagios for years and recently switched to Hobbit. Hobbit comes ready to graph the data it collects and data collected by your own scripts so you can see trends in server and service health over time for capacity planning, diagnosing problems, etc. It can be done in Nagios but its alot more painful to setup. I've found that Hobbit is capable of doing so much more then Nagios. Hobbit smokes Nagios. I used Nagios for years and recently switched to Hobbit. Hobbit comes ready to graph the data it collects and data collected by your own scripts so you can see trends in server and service health over time for capacity planning, diagnosing problems, etc. It can be done in Nagios but its alot more painful to setup. I’ve found that Hobbit is capable of doing so much more then Nagios.

]]> by: weismanm http://www.linux-mag.com/id/6138/#comment-1159 Fri, 06 Jun 2008 07:18:25 +0000 http://www.linux-mag.com/id/6138/#comment-1159 this looks very interesting, and I appreciate everyone's comments. I am always looking for easy to install and use systems and will try this. nagois, zenoss and BB are a bear to install and configure (at least for me) Has anyone ever used Big Sister? It's another BB clone, but quite easy to install and use, not so easy to configure past basic stuff, though. this looks very interesting, and I appreciate everyone’s comments. I am always looking for easy to install and use systems and will try this. nagois, zenoss and BB are a bear to install and configure (at least for me)

Has anyone ever used Big Sister? It’s another BB clone, but quite easy to install and use, not so easy to configure past basic stuff, though.

]]>