Docker: copying bootstrap data to pipe caused "write init-p: broken pipe"

[daath (Lars)]

Hi there,

I am currently out of the country, and silly as I am, I wanted to update my discourse server - usually that goes without issues.

After the update, the forum didn’t come back up. I tried git pull and launcher rebuild but this error keeps coming up:

/usr/bin/docker: Error response from daemon: OCI runtime create failed: container_linux.go:348: starting container process caused “process_linux.go:297: copying bootstrap data to pipe caused "write init-p: broken pipe"”: unknown.
Your Docker installation is not working correctly

See: Troubleshoot docker installation issues

Searching the forum didn’t yield much - other than “Your docker isn’t working”. I didn’t change a thing, and the machine itself is dedicated for Discourse. I tried reinstalling docker, but nothing changed…

Any help to get the forum up and running again will be greatly appreciated

It’s running Ubuntu 14.04.5 LTS on Digital Ocean (I wasn’t sure if it was safe to upgrade to 16.04)…

From the thread I ran:

root@forum:~# docker run --rm -i -a stdin -a stdout ubuntu echo 1
docker: Error response from daemon: OCI runtime create failed: container_linux.go:348: starting container process caused "process_linux.go:297: copying bootstrap data to pipe caused \"write init-p: broken pipe\"": unknown.

Docker version:

# docker -v
Docker version 18.06.2-ce, build 6d37f41

[Falco]

Did you try a reboot on the server?

[daath (Lars)]

Several, to no avail

[Falco]

If a simple

@daath

root@forum:~# docker run --rm -i -a stdin -a stdout ubuntu echo 1

fails, docker is completely broken in your server. Can you share the output of df -h ?

[daath (Lars)]

Sure. There should be enough space:

root@forum:~# df -h
Filesystem                 Size  Used Avail Use% Mounted on
udev                       3.9G  4.0K  3.9G   1% /dev
tmpfs                      799M  360K  799M   1% /run
/dev/disk/by-label/DOROOT  158G  117G   34G  78% /
none                       4.0K     0  4.0K   0% /sys/fs/cgroup
none                       5.0M     0  5.0M   0% /run/lock
none                       3.9G     0  3.9G   0% /run/shm
none                       100M     0  100M   0% /run/user

[Falco]

Is the docker service running? service docker start

[daath (Lars)]

Yep, running:

root@forum:/var/discourse# service docker start
start: Job is already running: docker
root@forum:/var/discourse# ps aux | grep docker
root      1687  0.2  0.9 881840 76188 ?        Ssl  13:26   0:05 /usr/bin/dockerd --raw-logs
root      1701  0.2  0.4 564568 35080 ?        Ssl  13:26   0:05 docker-containerd --config /var/run/docker/containerd/containerd.toml

[daath (Lars)]

I can’t figure it out. Could I create a new server (and use Ubuntu 18.04) and just install docker and copy over /var/discourse ?

[pfaffman (Jay Pfaffman)]

Yes. That should work.

[daath (Lars)]

I hope so. Currently copying over a crapton of data It’ll take a while

[pfaffman (Jay Pfaffman)]

If you’d be satisfied with the latest backup you could copy just that and the containers directory.

[daath (Lars)]

I need the latest of everything … It looks like it’s working - Waiting for DNS to propagate fully

[aye.javier (Javier)]

The problem is the version of the kernel and docker. It’s a security updete from docker

Docker Engine release notes

https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2019-5736

[daath (Lars)]

Man, I wish I had seen that… I spent 5 hours debugging, searching, then creating a new droplet and copying everything over and waiting for DNS to propagate Oh, well, now I have a 18.04 droplet running the forum…