Score:0

High interrupt and context switch count on Nomad / Docker

cn flag

We are seeing rather high interrupt and context switch counts on our Nomad clients:

# vmstat 1
procs -----------memory---------- ---swap-- -----io---- -system-- ------cpu----- r  b   swpd   free   buff  cache   si   so    bi    bo   in   cs us sy id wa st
6  0      0 317479392   4168 104093176    0    0     0     0 145232 306836 11  4 84  0  0

The average of ~150k in/s and ~300k cs/s are not spikes but sustained over the day. We analysed the processes (with pidstat), the Top 1 service type generates around 40k cs/s. We looked into networking as well, it seems most interrupts originate from there. So we argue that Docker networking is responsible for the high counts.

My Question: Are these expected values for a medium to large sized node? We are running on 32c/64t CPU (~20% busy), 512GB RAM (~140GB free). Thanks for sharing your experience.

mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.