0

I as looking through munin-generated charts on an Ubuntu 12.10 server.

I noticed that starting about 2 weeks ago(this server got a cold-shutdown at that moment) it started generating lots of rtc0 interrups. Anyone know what they are and why they might suddenly start appearing like this?

irqstats

quamis
  • 362
  • 7
  • 18

2 Answers2

1

rtc0 is a clock. You may have started a process that is using it, or a process may be using it. If you (for instance) read from it and it blocks, interrupts will be generated to signal the process that is blocking to read from it.

By no means does this indicate a problem. Hardware interrupts are pretty much how blocking I/O and timers work.

In general, it doesn't look like your system is spending a problematic amount of time servicing interrupts. Keep in mind that the "m" in the data under the graph is for "micro" (though "micro-interrupts per second" is awkward, it is a unit used in this data).

Falcon Momot
  • 25,244
  • 15
  • 63
  • 92
1

Please define lots. None of these graphs appear to be showing more than 1000 interrupts/second and mostly less than 100 i/s which quite frankly is a system ticking over.

From the image it's not entirely clear which is rtc-0 but it's either 1e-2/sec or 5e+00 sec which is nothing.

You are worrying over nothing or misinterpreting the data provided.

user9517
  • 115,471
  • 20
  • 215
  • 297
  • ok, i may worry for nothing, but why did it start out of nowhere? The only thing that happened was a cold-restart at that point. No config was changed, no errors were found, nothing... – quamis Sep 19 '13 at 06:54