What is most valuable?
- Multi-level infrastructure maps. You can set up highly customizable maps of your infrastructure, including each component status, maps which can contain submaps, and if something’s wrong at the submap level - the submap will be shown with problems.
- Agent and agentless monitoring. Installing an agent is the easiest way, but SSH, SNMP, Impi, custom scripting and other protocols are supported.
- Templates. There are a broad set of templates that define monitoring items for common OS and applications, you can build your own or use one developed by the community.
- An API for you to create, modify, and delete most things programmatically, including items, triggers, charts, etc.
How has it helped my organization?
As we know how are systems are supposed to work, Zabbix has given us the flexibility of building our checks without relying on special plugins. If some monitoring item is not provided already, building your own is very easy, and any scripting language will do. We’re able to pinpoint with accuracy where issues lie, and respond to them in a cost effective way, due to the nature of it being open source, with fairly decent documentation and optional commercial support.
What needs improvement?
The Java gateway for monitoring Java applications didn’t work for us. Our JMX Service URL’s are somewhat complex, but we were able to write our own JMX checks.
For how long have I used the solution?
What do I think about the stability of the solution?
When your checks grow to a certain volume, you need to adjust some settings on the server related to cache sizes, preforked workers, and so on. This is expected as with any service that grows beyond its default settings, there’s documentation about it.
What do I think about the scalability of the solution?
The system is as scalable as your database is. We applied partitioning on the biggest tables to keep the performant at 14 new values per second. You can setup a distributed architecture consisting of an N number of proxies capturing data and reporting to N master servers.
How are customer service and technical support?
Fortunately, we’ve never needed it, because we have very talented engineers in our team and there’s plenty of documentation available online. There are also user forums which are helpful most of the time.
Which solution did I use previously and why did I switch?
A fork of Nagios called Icinga, it’s pretty good, but something like the Zabbix Maps feature was a must have for us. We needed something very schematic and flexible to show the overall status of our systems.
How was the initial setup?
It wasn’t complex, but we have very skilled engineers.
The quickest way to go is to have a turnkey virtual instance tailored to your needs by Zabbix, or use one of the freely available ones.
What's my experience with pricing, setup cost, and licensing?
The product is open source with optional commercial support, which is recommended to reduce the learning curve, avoid pitfalls, and keep the project going.
Which other solutions did I evaluate?
SolarWinds. Nice product, but our particular project was more in line with the Linux, PostgreSQL, and the Open Source philosophy.
What other advice do I have?
If you’re into the Open Source and freedom principles, don’t mind getting your hands dirty, and have the technical skills, Zabbix is a good choice. Commercial support is available if you don’t want to go at it alone, lack sufficient technical skills, and you need help to keep the project going.
Which version of this solution are you currently using?