Best OSS Systems Mgmt App You Never Heard Of

FLOSSisnot4Teeth writes "You probably are familiar with Nagios and Webmin as two of the most widely deployed open source systems management applications. However, this month's SourceForge.net Project of the Month is probably a newcomer to open source systems and network administrators. Zenoss Core is a systems monitoring platform, released under GPL and over the last year it's become one of the most popular SF.net projects. Unlike most of these new "commercially backed" open source projects, Zenoss Core is the only version, their corporate sponsor doesn't offer a "pro version". Also their developers have been committing code back to other projects like RRDTool and Twisted. I have been playing around with Zenoss for about six months and have been totally impressed. Would be curious to see what other Slashdot readers think." SourceForge.net and Slashdot are both owned by OSTG.

Reminiscent of OpManager

[BinarySkies (920189)]

This seems a bit reminiscent of AdventNet's OpManager system. I would like to point out right now, though, that OpManager is about 700$ for a decent license that even compares to the kind of coverage you get from Zenoss. I wouldn't compare this app to Webmin so much; webmin controls only local system programs and some minimal enterprise software. This dives into the devices end of things as well, providing a decent number of MIBs. I'm very impressed by how the management console includes inventory on devices. Documentation seems decent, but then again I've been working with enterprise networking and systems management for several years. Even at that, this tool isn't demeaning to those who have prior experience. All and all a great OSS project and I look forward to seeing it continue to improve with time.

Re:

[muszek (882567)]

I just briefly looked at what Zenoss does... not what I was hoping for. I see a great need for FOSS web server management software. Some neat tools that would make configuring various services (apache, postfix, bind, vsftpd, sshd... stuff like that) easier for newbies. I've had my own dedicated server for a year now and still there are areas that I'm lacking at. I wanted to go 100% FOSS, so didn't get any proprietary control panel for that server. I'm using afraid.org, because I'm scared to touch BIND.

Re:

[jcam2 (248062)]

If you are looking for a tool to configure Apache, Webmin would be the way to go. One of its modules provides a relatively easy to use UI for editing httpd.conf and other Apache config files.

Plus there are other modules for servers like Sendmail, Postfix, SSHd, BIND, Squid and more..

Re:

[muszek (882567)]

Yeah, but Webmin doesn't quite make things easier. Looking at BIND module still makes me want to cry. It makes things easier for those that feel uncomfortable with CLI, though.

What I was talking about is a thing that kinda dumbs everything down, letting users set up the most common things. Example: in Apache few most common options + adding of virtual domains. In BIND, add domains, record type and target address.

Re:

[ahodgson (74077)]

What you're looking for is called a system administrator. Any server on the Internet requires one. Either you're willing to spend the time to learn how to be that person, or you should be paying someone else to do it. There is no piece of software that can do it for you. Believing there is will result in downtime and, sooner or later, someone hacking your server.

Re:

[jcam2 (248062)]

Yeah, as the developer I'd be the first to admit that Webmin's UI is still too complex - however, there is a tradeoff between functionality and simplicity, and I've generally gone towards functionality. Webmin isn't too useful if you are not familiar with concepts like zones and DNS records.. if you want something simpler, one of the hosted DNS services (network solutions, godaddy, etc..) would be better.

Documentation

[0racle (667029)]

There is no documentation listed on the Source forge site. Does it monitor Windows machines without having to install extra software on the Windows Servers and Desktops? Does it monitor routers and switches? Is there any documentation?

Re:Documentation

[PatMouser (1692)]

Take a look at http://www.zenoss.com/ [zenoss.com]. They just didn't link it in on the Source Forge site.

Re:

[Bios_Hakr (68586)]

My guess is that it uses SNMP to monitor systems. SNMP is included with every version of Windows from NT on up that tree. I think you have to install the service separately and then enable it with the proper community strings. After that, you can track memory, HD, processor usage. We used to use it with MRTG and HP Openview to find out which users were saving shit to their desktops vice the network server.

Re:Documentation

[bastion_xx (233612)]

Besides the python script to create a WMI monitor, zenoss does have good SNMP support - when you run the Informant http://www.wtcs.org/informant/ [wtcs.org] SNMP extensions.

Re:

[nbannerman (974715)]

Sources for documentation from their main website;

http://zenoss.com/docs/zenwin [zenoss.com] - Windows documentation, rather brief. Supports 2003/XP apparently.
http://zenoss.com/docs [zenoss.com] - Main documentation website for Linux / BSD.

Re:

[linux_warp (187395)]

Check out: http://zenoss.com/docs [zenoss.com]

Re:

[bastion_xx (233612)]

zenwin handles Windows boxes using WMI. Cool part is that only one system needs to go through a very convoluted install to then query any other servers in the same domain.

Monitor and switch support is good for common devices. I've run into problems trying to monitor things such as m0n0wall devices (it tries to pull CPU and memory stats, but the OIDs are not responding correctly). Cacti seems to do better on this right now, but the alerting from Zenoss makes monitoring much easer.

Zenoss has made great stride

Re:

[Bandman (86149)]

Me too, but lord I hope the config on Zenoss is better than on nagios

(here, better == easier)

Re:

[Feyr (449684)]

i've tried the VM app of zenoss. the config is easier at first glance, but it gets tedious as soon as you want to monitor something other than the default.

Re:

[Cylix (55374)]

For nagios, I setup nagiosql at least I believe that was the name of the project.

I don't believe it's under development anymore, but a little tweaking on the sql side and it was working fine for me.

After that, it's pretty much a breeze to configure and add servers. Once I wrote a custom script (many were provided) it could be applied to any box.

The only drawback is all configuration files are kept in sql and regenerated. However, if you are using it to edit your configs... chances are you may not want to tw

Re:

[Anarke_Incarnate (733529)]

Hyperic HQ seems to beat this. ZenOSS is good if you want agentless (except for Windows) monitoring or need a virtual appliance. You can build Hyperic up on a Virtual Machine.

Look at hyperic.com
I don't work for them, just spent months looking at a solution for this.

Re:

[Bandman (86149)]

Yea, it's very flexible and powerful, but it's also a complete pain in the ass, especially for adding several service checks that are very similar.

I have a

check_disk_root, check_disk_var, check_disk_db, etc etc

and each one requires an entry in checkcommands and services.

Re:

[rhizome (115711)]

and each one requires an entry in checkcommands and services.

You should be using "servicegroup," then.

Also, you might like http://fruity.sourceforge.net/ [sourceforge.net]

Re:Documentation

[bastion_xx (233612)]

Don't get me wrong, Nagios rocks and has been a godsend, even back in the netsaint era. Where Zenoss is useful is adding new devices for templates that already exist. In Nagios it's either change the underlying config files, pre-flight, and reload, or use a GUI to do the same.

Our use has transitioned from hand-crafted nagios plugins for bespoke services to more generic checks and longer term capacity planning. Zenoss can do this, and it appears with less operational management, allowing us to focus on performance data and more in-depth Windows monitors (again an internal change from Linux core systems to Windows -- different client base).

Which Month is Right?

[by Anonymous Coward]

http://sourceforge.net/projects/zenoss [sourceforge.net] says:
Project of the month for : February 2007

http://sourceforge.net/potm/potm-2007-03.php [sourceforge.net] says:
Project of the Month: March 2007 - Zenoss Core

Looks like a newcomer alright...

taken from greg haywoods blog

[by Anonymous Coward]

Zenoss
Jan 26, 2007

I may have finally found the perfect monitor solution for my network: Zenoss. I have been using Nagios + Cacti + Smokeping for quite a while now. It works, but it's not integrated, and for many services, I'm running 2-3 checks. Running those every 5-10 minutes generates a tremendous amount of traffic (during the last 2 weeks, the monitor station has caused 20% of all traffic crossing the primary firewall!). The closest all-in-one I'd found previously was OpenNMS, which is so difficult to really understand and manage well, and so didn't fit my needs. I'd given some thought to rolling my own in Ruby, but just don't have the time for such an undertaking.

So while browsing the rPath/rBuilder site this morning, I discovered Zenoss. It's Zope-based, which I find a bit interesting. But from what I've seen in the 30 minutes I've had it running, the developers are right on with what I've been looking for. It has auto-discovery support, placing everything into a "/Discovered" group if it can't pick the right group on its own (the firewall was placed into the "/Network/Routers" group since it was part of the discovery chain). But it is smart enough to correlate different IPs to a single device, which OpenNMS can't do. It also supports Nagios plugins (though only via ssh and not nrpe), so I can leverage that investment while I evaluate the Zenoss way of checking.

There's also a built-in syslog catcher, so it can correlate log events to devices, which could be another huge time saver. And it has asset/inventory management so I don't need to keep that data separately either. What can't this puppy do?!

You can install from source or RPM, and there's a vmware image available too. It requires Python 2.3.5+ and MySQL 5.0.22+. Since I wanted to run on my Debian Sarge monitor station (which already has access to all the devices to manage), I had to upgrade the DB. Easy enough with the backports. The only trick I ran into there is that the install process requires port 8100 be available. You can change after install, but I couldn't find a way to change prior. The installer doesn't notice if the port is already in use, it just silently fails, and so when starting the Zope DB setup, it gets in a loop of printing "." (dots). Finally realized I had to shut down a Mongrel-run Rails app to get it going, and it worked perfectly. (Bug #933 has been filed.)

Stay tuned for more, as I will be playing with this ALOT over the next few weeks!

Litmus test - Failed

[Dareth (47614)]

You have searched for packages that names contain zenoss in all suites, all sections, and all architectures.

Can't find that package.

----

It appears I am not yet interested.

Re:Another litmus test

[Jason Earl (1894)]

I am as big a PostgreSQL bigot as you are likely to find, but I don't see the problem with using MySQL for storing monitoring data. I mean seriously, why should I care if the application stores the fact that my servers still respond to pings in a transaction safe manner? Nagios, which I currently use, stores this information in flat text files.

Advertise on Slashdot.org -- It's Free!

[phish (46788)]

Interesting how the submitter writes the post suggesting as if they're a user....

"I've been playing around with it for six months and have been totally impressed!"

Easy to be impressed by your own products, isn't it?

Isn't this a dupe?

[jkrise (535370)]

I can't understand why this isn't tagged dupe already... I seem to remember ZenOSS on /. a month or so ago... followed by an article on OpenNMS as well.

http://slashdot.org/article.pl?sid=07/02/23/003223 3 [slashdot.org]

I think Vista has broken most commercial network mgmnt offerings... nothing else can explain these dupes!

other contenders

[OriginalArlen (726444)]

As it happens I was just reading my locally saved copy of this related Slashdot piece, on OpenNMS [slashdot.org]. Other alternatives mentioned in the comments were:

• Cacti [cacti.net] (an RRDtool front-end -- if you don't know what RDDtool is, you don't need this :) )
• Munin [linpro.no], and
• OSSEC [ossec.net].

I've looked over someone's shoulder at the latter - it seems pretty good, it runs on SNMP - I tinkered with NAGIOS five years ago and found it good, but a little dangerous if you didn't read the docs before firing it up (back then, anyway, it auto-discovered the local network by strobing everything in sight with Nmap scans)... but I've no experience of any of these in production. I've been asked to build out a new office network, which will be a template for future local offices, and getting the monitoring right is going to be crucial, so any actual experience of production use gratefully received!

Re:other contenders

[fimbulvetr (598306)]

You forgot hobbit. Which is a gpl version of bb, written in c (much, much, much faster). Support for clients on servers (so you don't have nasty snmp everywhere). Integrated rrd, "smart" checks (i.e. looks for status 200 w/ http, looks for +OK on tcp/110 connections, etc. It doesn't just check that a port is open).

Project's demo:

http://www.hswn.dk/hobbit/ [www.hswn.dk]

Re:

[OriginalArlen (726444)]

I hadn't heard of it; thanks!

BTW -- "bb" == "big brother"? I haven't looked at that much, but I've seen the agent UI and I nearly threw up from the vertigo of being flung back in time to 1996 and VB5. Egad, it was like when desktop published first took off and lo! the departmental newsletters were many, and terrible ;)

Re:

[fimbulvetr (598306)]

Yeah, big brother. I like it because it uses tcp port 1984. Very funny. Anyway, the interface isn't beautiful, but that's not what guys like me are after. With just one page I can quickly ascertain the status of my network as a whole. Depending on how you lay it out, you can have groups/groups of groups or everything on one page.

I like it mostly because of the speed. With several thousand servers, nagios and many other snmp-like monitoring tools start dying a horrible death unless you do things spectacularl

ZABBIX

[by Anonymous Coward]

Did you ever heard of ZABBIX [zabbix.com]? I believe this is the best Open Source monitoring solution around. It is a mature and flexible piece of software which comes with very impressive feature set.

Re:

[zeenixus (571630)]

it's also not a huge gob of python like zenoss which requires at least 512 megs of ram to run and do something useful.

zabbix server and clients are written in C, with php for the web front-end. It has it's quirks and shortcomings, but I'll take simple and lightweight over bloaty probably-can't-scale-worth-a-damn any day.

Re:

[Anarke_Incarnate (733529)]

How does it compare to Hyperic HQ? I have not heard of Zabbix but its pretty and looks interesting. It supports HP-UX, which is a plus here. Most of the monitoring solutions I have found support Win/Lin only

Nice looking, but not enough to meet my needs.

[Harker (96598)]

The company I work for is currently looking into remote monitoring of the data center environmental conditions as well as server status. So far, I've found very few options that will do this, and none seem good enough (one will monitor one of our UPS's but not the other one for example).

We currently have one web-based monitoring tool in place for server status, and I doubt they'll be willing to change to another, especially if it is open source. The last time I mentioned an open-source alternative (change

Re:

[ErnieD (19277)]

If you don't already have all of the monitoring hardware in place, check out the NetBotz (now owned by APC) suite of monitoring products. You buy one rack-mount system, and can tie any number of sensor "pods" (or even third-party sensors via customizable input contacts), and all sensors report back to the main unit for logging and alerting. It can also forward the SNMP traps onto an existing network monitoring system if you have one.

I don't believe they'll monitor UPS equipment however. For that, here at my

Re:

[Harker (96598)]

I'm not really certain what we currently have. I know that the floor (water) sensors are minimal, only existing right near the cooling units. They will need to be expanded at the very least. I'm figuring that if we have to expand them, then replacing the few that are already in place won't be that big a deal.

Beyond that, I really don't know what, or even IF we have anything except server monitoring. Currently, we depend on hourly walkthroughs of the data center...

I'll check into this as well though. Th

Re:

[loginx (586174)]

I'm pretty sure it's possible to accomplish these tasks using Open-Source software, however, if your management isn't open to the idea, there is nothing wrong in buying a commercial product.

I work for a Canadian company called Netmon [netmon.ca] and we sell both the network monitoring product (Netmon) and the environmental probes that send info about humidity, temperature, etc... right onto it. It is also able to monitor UPS's, Cisco gear, etc... and runs on a Linux server.

I know I'm shamelessly plugging our company

Re:

[Harker (96598)]

I'll add that to my list of services to look into. If we can find an all-in-one solution, or one that can be built up from scratch given what little we actually have in place now.

Not certain if they'll like the Linux server bit though. We have only one Unix admin, so that might be an issue. I'm willing to learn and add to my linux knowledge though...

H.

Re:

[loginx (586174)]

H.
There is no administration required. We take you through the initial set up and your box fetches updates automatically. Everything else is done through a web-interface, and your initial contract includes a full year of complete support for the appliance, so we can walk you through any task. If you open up SSH access to our private IP range, we can also log in and help you out with anything you're trying to do with it. The upcoming version has a lightweight graphical desktop used for deployment (setting up

Re:

[Harker (96598)]

I'll keep that in mind. I should point out, in case I haven't already, that I'm doing this on my own, to kind of push the project onward, and to make myself look a bit better in the eyes of my superiors. My direct manager is in charge of the overall project.

I will look more into this later this week though. The bookmarks are already down.

H.

It's pretty good!

[wireloose (759042)]

I'm using Nagios on a group of core network devices. I have to be super careful not to perform an "up2date" on it (using RH 4.0 EL) because the Nagios packages always overwrite my config files. But Nagios is good, and it's been very useful. It took me a few days of work to get it set up the way I wanted, and it's been a charm ever since.

A few weeks ago someone posted an article on the top ten OSS projects to watch, and Zenoss was one of those projects. I downloaded it to experiment. I had it up and running in about 20 minutes, on Ubuntu. It's far more powerful at its ability to gather data from nodes. And setup is far less manual. Network discovery worked very well. It found devices on our network that we didn't know were out there. It required no integration with other packages. The interface is also more intuitive in some areas, such as viewing event histories. But, it's more challenging to find performance charts the first few times.

I especially liked the automatic snmp walk through the MIBS on each device. This makes it much easier to pull statistics from it, without having to edit text files. The MRTG-style charts are also good. I wish they were more readily configurable. I also wish there were more MIBS in the distribution, but you can find most by carefully searching equipment provider's web sites.

All in all, After running it side by side with my Nagios setup for a couple of weeks, I like it much better. And I'm moving more SNMP agents into my network just because of Zenoss.

Re:

[morgan_greywolf (835522)]

It may very well be a good package, but until such time they use PNG for their screenshots, going through the tour involves squinting hard. Unfortunately no schooners appear, begging the question why JPEG was used to start with.

I'll ignore your blatant misuse of the phrase 'begging the question', but answer your specific question: JPEG was probably chosen because not all browsers support PNG -- especially true of older versions of IE -- but, they do all support JPEG.

Re:

[Lord Ender (156273)]

That doesn't matter. If he wanted the 12 people still using IE4 to see his site, he could have used GIF.

Attention everyone: Lossy image compression, such as JPG, should NEVER EVER EVER be used for things with solid colors or high contrasts (especially text)! Use PNG, GIF, BMP, or any other losslessly/un-compressed format. JPG is only for "natural" images such as photographs!

Compressing text with lossy image compression is a sin almost as bad as butchering "beg the question." BOTH of these atrocities make th

Re:

[drinkypoo (153816)]

Yes, no doubt an Open Source system management application is terribly worried about supporting all those IE 3 users out there who want to look at the screenshots.

We're talking about a Open/Free Unix-based management system. They're probably worried about supporting Mosaic.

Re:

[morgan_greywolf (835522)]

Actually, Netscape Navigator didn't have PNG support until version 4.04. And PNG support in IE browsers 4.0-6.0 has been abysmal. IE 4 will crash on any PNG with metadata, and IE 5-7 have problems with proper gamma support. And IE 4-6 doesn't support PNG transparency, at least not in anyway that's useful.

Re:

[drinkypoo (153816)]

IE6 supports PNG transparency through the AlphaImageLoader ActiveX control, but my experience has been that it does not load reliably, so yes, you're quite right. I just wanted to reply and say this stuff before someone else showed up and said "What about AlphaImageLoader?" For me, it would work on something like 50% of page loads.

Re:Here's what I think

[mabinogi (74033)]

Slashdot is for anything the Slashdot staff feel like approving.
It's not for satisfying anonymous cowards.

Re:

[just_another_sean (919159)]

Not to mention with the advent of the Firehose I get to be a part of the
selection process now as well.

Re:

[grimwell (141031)]

You laugh but I had a job where ping was their "solution". First day of "training" they show me this batch file that I needed to run in the morning. The batch went and pinged everything, if it ran successfully then everything was up and life was good.

Yeah, that was an interesting gig.

Re:

[firl (907479)]

Incase your wondering he said support, not purchase.