Icinga vs. Sensu, by Sumo Logic

Icinga is a world-class monitoring system. It can be used for most general monitoring situations. It is not a silver bullet, however, and there are instances where domain-specific monitoring systems are necessary. However, the output from those monitoring systems can be funneled into Icinga as a central monitoring and alerting system.

Incentivized

• Well suited for raising pagers when you have count-based metrics.
• Well suited for threshold-based metrics.
• Not well suited for tracking events over the day, since graphs and other visualizations are missing.

Incentivized

• Wealth of community-developed plugins.
• Stable codebase.
• Icinga 2 supports distributed monitoring.
• Very performant, can support tens of thousands of checks per server.

Incentivized

• Unique concept as a "monitoring router" that can tie services together.
• Backward-compatible with Nagios environments.
• More of a DevOps focus than Nagios.

Incentivized

• High learning curve, setting up Icinga from scratch can be a bit of a challenge starting out.
• If the io2db process fails you UI stops updating, which can be very frustrating.
• There is no simple mechanism for adding new hosts and services through the web UI, it's all very config-file based.

Incentivized

• Cannot be solely used for tracking metrics over time.
• Not a very good UI.

Incentivized

Icinga is a solid solution which does everything it promises. It is backwards compatible with most Nagios instances, making the transition very easy. Once you get the hang of installing new plugins and editing configuration files expanding its monitoring capabilities are easy.

Incentivized

Sensu's customer support was always willing to work with us but never really seemed to learn much from our experiences. I think they get a lot of customers with DevOps IT teams that are willing to put in a lot of elbow grease to get the most of Sensu's architecture. However, despite explaining my continued disappointment with their documentation and the overall flow of the product, I never got much more than a "sorry" and a notice that their documentation was open source if I wanted to contribute to it. The problem, of course, is that you can't document what you don't understand. I'm a former technical writer, so I know that better than most.

Incentivized

Icinga is better than Nagios because of its nicer user interface. New Relic can monitor CPU/memory and disk usage, but it's more of a performance and application troubleshooting tool rather than monitoring

Incentivized

Have used New Relic and Sematext Cloud for APM and for tracking over days and visualizing the issues. But those are very expensive as compared to Sensu.

Incentivized

• With one check you know which applications are faulty e.g. after an upgrade. Which is big time saver
• You easily detect outages ion the applications so that your customer ideally does not even realize there was an outage.
• Detect if the environment does deliver the same result as in the same time as before to detect shortages.
• Additional information when debugging. Saved us several hours where we could simply point to a database which was slow.

Incentivized

• Standing up the Sensu Go server took very little effort.
• Setting up and maintaining the build processes and deployment logic for Sensu assets and checks was somewhat exhausting and resulted in lower adoption among non-DevOps IT.
• The limited web interface resulted in lower adoption among non-DevOps IT.

Incentivized