search

firecracker-microvm / firecracker-containerd

firecracker-containerd enables containerd to manage containers as Firecracker microVMs
Apache License 2.0
2.22k stars 182 forks source link

Cleanup logs around io streams shutting down #367

Open sipsma opened 4 years ago

sipsma commented 4 years ago

Right now when a container exits you get some confusing error logs like this:

time="2019-12-11T20:03:38.436612731Z" level=error msg="error copying io" ExecID= TaskID=container-80-3 error="read unix @->firecracker.vsock: use of closed network connection" runtime=aws.firecracker stream=stdout vmID=80

or this:

time="2019-12-11T20:03:38.443037999Z" level=debug msg="[   81.326403] agent[848]: time=\"2019-12-11T20:03:38Z\" level=error msg=\"error copying io\" ExecID= TaskID=container-80-3 error=\"read /proc/self/fd/76: file already closed\" stream=stderr" jailer=noop runtime=aws.firecracker vmID=80 vmm_stream=stdout

I don't think these in particular should be logged at ERROR level and they should probably be clarified as they just indicate the stream got closed. Something more like std<in,out,err> closed at INFO level seems more appropriate. We can include the full actual error message at DEBUG still. We'll also want to make sure that if other unexpected errors occur, we continue to log those at ERROR.

Zyqsempai commented 4 years ago

I can take this thingy;)

sipsma commented 4 years ago

Sounds good, thanks @Zyqsempai !

acpana commented 3 years ago

@kzys, can we close this given https://github.com/firecracker-microvm/firecracker-containerd/pull/369 ?

kzys commented 3 years ago

I want to have some tests to prevent regressions.