Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Our development server is running debian 4.0 (etch) as a guest under KVM. For the longest time, it had sporadic issues where many processes would appear to just hang. We finally found the issue - the clocksource was still configured as tsc from when it was running on vmware server. I switched it to "pit" and the problems disappeared.

Unfortunately, I just discovered that under high cpu and high disk load (large number of selects/updates on a locally hosted db), we start to see clock skew on the guest. When running under high disk load and moderaly high cpu load for half an hour, we ended up with a clock skew of about 200 seconds. Just this morning I noticed a 7 second skew.

Is this related to the clocksource, and if so, which is a better choice? I haven't been able to find a lot of information online explaining the difference between each clock source.

cat /sys/devices/system/clocksource/clocksource0/available_clocksource reports:

jiffies hpet tsc pit

Due to supporting our legacy applications, I cannot upgrade this server to a newer version that includes kvm_clock.

share|improve this question

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.