Join GitHub today
GitHub is home to over 50 million developers working together to host and review code, manage projects, and build software together.Sign up
Fleet can not get response from etcd and CPU utilization is very high #1119
I set up a CoreOS cluster (3 VMs) on top of my OpenStack environment, and I found etcd seems running well (I tuned etcd with "peer-election-timeout: 3000" and "peer-heartbeat-interval: 600" to avoid heartbeat timeout:):
But it seems fleet is not running well:
And I found the CPU utilization of my CoreOS VMs is very high:
And "fleet list-machines" can not list machines:
I am using 557.1.0:
Sorry, my bad, the output of
and just now, I added the following in my cloud-config:
And then it seems fleet can get response from etcd now, but it has other errors:
Are these errors harmful? And I still see the CPU utilization of fleet is very high:
The logs of update-engine and fleet are too big, and it seems I can not attach files here ...
Currently in my CoreOS VMs, I still see the CPU utilization of fleet and etcd are very high but not update-engine, and there are a lot of error in fleet log. However, it seems these error can be ignored, so I do not know why fleet is so busy.
and for etcd, its log seems OK:
So I do not know why etcd is so busy too.
And it seems my CoreOS VMs have been automatically updated from 557.1.0 to 557.2.0.