Skip navigation.

Comments on Society, Technology, and Life

Posts tagged with "IT"

Reviewing the Zenoss Book from PACKT Publishing

, , , ...

The book Zenoss Core by Michael Badger is a professional introduction to the Zenoss Core Open Source network monitoring system. For those who don't know, a part of being a System and or Network Admin is knowing whether your servers and switches are working. Checking every system manually was impossible, and waiting for users to submit a trouble ticket is less than ideal.

Traditionally there has been a mishmash of available tools, both commercial (HP Openview etc) and free and Open Source (Nagios etc) to help with this task. They can check aspects of your services to see if they are working as expected, or are down. When there is a problem they can change a Web GUI indicator, send an E-mail or take automated corrective action. The problem had been that the commercial tools were quite expensive, and out of the reach of smaller businesses. They also had limited ability to customize them, and you were of course license constrained, even in large businesses. The Free and Open source tools took lots of time and experience to configure, and support options were varied. This made some Enterprises skittish about choosing them. Zenoss is a great middle ground, there is an Open Source version that pre packages lots of the OSS tools, ready to do many tasks out of the box. There is all the extensibility that access to the source code, as well as defined APIs provides. Then there is the Enterprise version which comes with support from the developers as well as pre made plug-ins that do even more than just the Core version.

This book is a great compliment to the official documentation. It's sort of a "missing manual" to Zenoss. The book is written by a professional author, and it shows. It's laid out clearly and takes you through the GUI step by step. A user who isn't an admin could pick up this book and be understanding alerts and where to find graphs in a day. This fills an important role for just using Zenoss that the Admin and Developer guides available on the web don't. The on-line documentation often assumes you know what the jargon means or where the "Event Console" is. This book does not. I wish I'd had it before I started trying to learn the product. As a base, it makes the Admin guide much more understandable.

It's more than just a users guide though. It goes over such intricate topics as event mapping with more than just examples, it actually explains what the boxes are that you're typing into. It explains alerting and other parts step by step. It's an indispensable reference for Zenoss users and Administrators. Beginners should get this before wading into the forums or Admin guide. It will save you time.

Each chapter tackles a major section of Zenoss. It starts with the basic GUI aspect, moves on to any details needed to actually use it with your equipment. There are usually several sections where input boxes are explained, use cases discussed and examples presented. Finally there is a summary of the chapter to synthesize the entire concept for you.

The two appendixes are what Zenoss Masters want though. A short 10 pages so, they list the missing event fields used in Event mapping, and many of the variables used in the TALES expressions that are available no where else, save possibly strewn throughout 2 years of forum entries. If you are trying to move beyond the out of the box experience with Zenoss, you'll want it for just the appendix. The rest will just save your Zenoss Guru time training the end user admins.

The book isn't perfect though. Some places it is too terse. In others it could use an example in addition to just telling the reader what a box is. The book is a little short for the price, and there's a lot of places in Zenoss where more could be said. I sense a second book that goes into more development detail, more of the commands or custom data sources you can create.

That said, the book is exactly what it says it is. A step by step guide to installing and using what's there aimed at new users. It also makes a good reference book. I especially use the Event chapter, and in the second edition would like to see that expanded.

You can get more info on the book here:
http://www.packtpub.com/zenoss-core-network-and-system-monitoring/book

Free Full Disk Encryption

, ,

I already posted this in the software forum, but I thought I'd link the discussion from here as this is really exciting IMO.

http://my.opera.com/community/forums/findpost.pl?id=2094777

Software for IT Management

, ,

Now that I am working in IT managing a diverse network of devices, I've aquired some experiance with what might work and might not work across those devices. One thing I've noticed is you can spend a LOT of money on various multi access tools to get some idea of what's going on with your Windows, Linux, and MacOSX machines, plus you might have Solaris, AIX, HP-UX or BSD machines. Fun.

Now, I haven't had the ability to compare various commercial offerings beyond looking at their documentation as each is quite expensive. But for those of you without oodles of dollars, I have identified, and used with some success some FLOSS tools. Some of these are really quite mature, some are usable with great potential. For some reason, many of these seem to come from France.

I could talk for some time on various tools, but for now I'll just lay out what I was looking for, and what I've (well, the team I'm on) have done.

We wanted something that would work across Linux, Windows and MacOSX. More is great, but those were important - we wanted to consolidate our management tools, not increase them. We already have management for each platform separately except for inventory of some sort, so that's were we started - inventory. And there is a great package of two projects that does this nicely, and has a useful package deployment tool as well.

OCSNG + GLPI. The "names" for the two projects.
ocsinventory-ng.org
glpi-project.org
So, you have to have OCSNG - you can layer GLPI on top if you want. OCSNG has clients that run on your machines, and check back to the server every so often, we use 1hr intervals. This isn't a problem as we only have it do stuff every 24 hrs, and you could of course increase this if you wanted.

With OCSNG you get a great database of what's on your machines - what hardware, OS, service pack, software, and even username (if someone is logged in when the inventory runs). So right away you have all this info programatically gathered. It also supports deploying software to Windows and Linux (though we only use it on Windows right now for that).

The nice thing about it's deployment technology is it's a pull from the client. So no firewall issues or insecurity on the remote machines. They use SSL to authenticate the commands, though the download is in the clear. And it works great on laptops, because if the client isn't on the network when you set the deployment, no problem. It just picks it up the next time it's on and starts the install. Same for desktops if users have incorrectly turned them off at night.

This is great for a current snapshot of the machine, but what if you want to track changes? GLPI is the answer - it integrates with OCSNG and reads in the database every 5 mintues or so. GLPI also supports software license tracking, network layout modeling (though this is entirely manual)- help desk ticket tracking + knowledge base/FAQ and plugins for more.

GLPI isn't as mature as OCSNG (which is at v3.0.1 or so even if it's NG v1.0.2 or so), it's at 0.68.3 - but mostly it's entirely useable for hardware/software inventory, location etc tracking. I expect you could use it as a helpdesk ticket tracking as well, but we already use RT so...

Finally, you might say - this is great, I know what I've got out there, where it is (with some data entry) and I can push software. How do I track if it's actually on in near real time though? What do I do with my switches etc?

Well, the answer is ZenOSS Core.
Zenoss.com

Pretty easy setup, great community support, and full commercial support if you really want to spend money. This does full SNMP monitoring of v1 & v2 SNMP. But so what, lots do SNMP... It also supports WMI monitoring, so the same box can monitor our Windows Services! This is big. Plus, it does nice graphing for performance monitoring if you use the free SNMP Informant on Windows, and set up your permissions on your linux boxes. It scales to 4,000+ machines if you've got the RAM. It supports multiple distributed monitors(not that you'd need that unless you have a truly huge network).

It does full e-mail alerting + event reaction, that is you can have it fire off an e-mail, and have it restart the service remotely + e-mail you the problem has already been fixed! You can build really complex alerts and hirearchies of notifications if you want.

If you're looking to update or possibly change how your deployment and monitoring works, FLOSS is looking more and more attractive with the responsive communities, open standards and of course free and commercial pricing and support.
December 2009
S M T W T F S
November 2009January 2010
1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30 31