This article is being updated. Please be aware the content herein, not limited to version numbers and slight syntax changes, may not match the output from the most recent versions of Bright. This notation will be removed when the content has been updated.
Good admins itch to wring every last drop of possible performance out of a cluster, so they naturally want to get rid of fluff that slows it down. Which brings up the question: what resources does the backend core of Bright Cluster Manager — that is CMDaemon — use?
Very little, it turns out. The CMDaemon application was designed and crafted with care, so it runs pretty close to the bare metal, with very little waste considering what it does.
Figures, from the designer:
On compute nodes, CMDaemon uses about 66MB of resident memory.
With the default monitoring configuration it uses about 17.5 cpu-core seconds per wall-clock day (measured on Intel Xeon L5320 1.86Ghz).
On a 16 core machine, this means CMDaemon consumes about 0.00126% of your CPU cycles. Over a 5 year period this means you lose about 33m of compute time.
If your own cluster seems to be way off the kind of figures you expect from this, taking into account the kind of load you place on CMDaemon for your usage and hardware, then you should certainly contact us to check if you have configured things correctly. CMDaemon is the cleanest and most efficient cluster manager in existence for the features it provides.