Is performance that great?
Yes it is. Nagios® already has fair performance, but Shinken is even faster than Nagios®. And it’s no surprise : Shinken was born as a proof of concept to remove all bottlenecks from the legacy Nagios® core.
Shinken uses advanced programming features like a process Pool, in memory data pipeline, mature distributed object model using Pyro, modular interfaces, cross-platform compatibility and a modern efficient language.
With standard Nagios® you can monitor up to a 1000 servers with a single server (with an average of 10 service checks every 5 minutes per server), and maybe 3000 with advanced tuning. If you manage to add high availability or database export, this value will fall.
With Shinken, you can easily have up to 10000 servers with one box, and of course high availability and database export! If you need even more performance, you can add daemons as resources on other servers : Shinken will automatically manage the configuration for the number of daemons you have!
Shinken can scale out horizontally on multiple servers or vertically with more powerful hardware. Shinken deals automatically with distributed status retention. There is also no need to use external clustering or HA solutions. Shinken servers can be in different IP subnets which makes it easy to deploy in two data centers or in a Hot disaster recovery site.
Flexibility for small installations
If you have less than 1000 servers, is there a benefit to go from Nagios to Shinken? YES, use less server resources, install it on linux and windows to exploit native capabilities.
Is your Nagios server a physical one? Why? You may need a physical server for your Nagios, but with Shinken, you can move to a virtual machine! Once on a virtual server you can cut down on required memory and processing by using a Linux distribution that supports JEOS (Just Enough OS), like Ubuntu Server edition.
It may also be that you have little hardware and would need to run it on a eeepc.
You can see below a picture of Shinken running with a 1000 server configuration on a eeepc 701 (proc@660Mhz). Such a configuration was pushing the limits of a Xeon@3Ghz Nagios® server.
Availability supervision AND performance data?
Shinken can handle performance data returned by checks seamlessly. There is little overhead, all operations are in memory and streamlined through Shinken’s cloud-like architecture.
Performance data is extracted from check results in broker modules responsible for formatting and forwarding the data to the Time-series database systems. Data is stored in an external database and Shinken’s web interface has builtin access to display the data using powerful graphical views.
The following time-series storage systems are supported:
- RRDtool: The graphical viewing plugin is called PNP4Nagios
- Graphite: The Shinken WebUI has native support for graph templates and seamless data access
PNP4Nagios is the mainstay, but Graphite is fast becoming THE preferred method of doing statistical analysis and graphical output for time-series.
Looking for other features?
Huge performance is just one feature of Shinken. Go to the feature list.