I've experimented with the cause/effect relationship of the NDO utils running as the event broker, and I can say positively that it is "causing" my service graphs to have lots of gaps for checks that write metrics to RRD files.
I'm honestly at a loss as to why, though. The load hasn't gone up on the nagios server (if anything, it's gone down a bit). The MySQL backend is very comfy (low load, again). All rrd writes are done *within* the plugin (both with an apan wrapper and custom scripts). Yet, the checks appear to be happening (alerts are still going out, etc). Definitely perplexing. Any ideas?
signature.asc
Description: This is a digitally signed message part
