Closed myugan closed 3 years ago
@myugan, we fixed an issue a few weeks ago that would explain what you describe above. Can you please try with our latest code by building from sources?
Okay got it, is it possible to create deb package for the latest update instead of compiling from the source.
At the moment you cannot create deb images by your own. However, we are about to start our next release cycle, so we are expecting to have the new images ready soon (one or two weeks).
As explained above, problem has been already fixed in top-of-tree. Please let us know if have any other question on this matter. Will close this one now.
I want to ask, there is any possibility the docker container with dead
state is caused by sysbox
?
Hi @myugan:
I want to ask, there is any possibility the docker container with dead state is caused by sysbox?
I've never come across that. In what context are you seeing this?
I'm not sure about this but some time in our system there is a container stuck and need to restart the daemon, when I see the process with docker ps -a
it turns to dead
state
I'm not sure about this but some time in our system there is a container stuck and need to restart the daemon, when I see the process with
docker ps -a
it turns todead
state
Got it. The error you reported above could lead to a container getting stuck / hanged.
As @rodnymolina mentioned, this error should be fixed in Sysbox top-of-tree. Have you seen this error with the Sysbox top-of-tree?
Ah okay, so that's the case because I still use the latest update deb package in Ubuntu and still waiting a new latest update as @rodnymolina told me. Thanks in advance
This is related to sysbox? like i said before, the docker is stuck sometimes when i remove the container.
cc @rodnymolina @ctalledo
Hi @myugan:
This is related to sysbox? like i said before, the docker is stuck sometimes when i remove the container.
Need more info to help you here.
What process is that strace from?
I suspect it's the docker CLI (since it's trying to open /var/run/docker.sock
). I don't see how Sysbox could have any effect on the Docker CLI communicating with the daemon.
Hi @ctalledo seems this is not coming again, you're right this issue doesn't come from sysbox itself but another configuration in docker that leads to stdout being buffered
so its turn the container stuck.
Hi @ctalledo seems this is not coming again, you're right this issue doesn't come from sysbox itself but another configuration in docker that leads to
stdout being buffered
so its turn the container stuck.
Thanks @myugan for root causing and confirming that it's not a Sysbox related issue, much appreciated.
Hi @ctalledo
Now I have the same issue but this is not coming from the Docker configuration, these things only happen sometimes that causes the container in Created state for a minute then the container can be running (Up state)
Unexpected error during NotifReceive() execution (bad file descriptor) on fd 17 pid 9813
Thanks @myugan for the update.
Is this behavior reproducible? If so, what are the steps to reproduce it?
The Unexpected error during NotifReceive() execution
warning is usually not a problem, it sometimes occurs when a process inside a container performs an system call trapped by Sysbox (e.g., mount syscall, etc) but then dies unexpectedly (usually because the container associated with that process was stopped).
Thanks as always for reporting.
Unfortunately, this is not reproducible at the moment but I will let you know once I can find a way how to get this issue consistently.
Hi @ctalledo @rodnymolina
I have an error log like this, i'm not sure because of this our containers sometimes crash can't kill the running containers without restart the docker daemon.