Docker Community Forums

Share and learn in the Docker community.

Huge image log issue


(Troff) #1

Hi!
Ubuntu 16.04, Docker-engine 1.13.1
i have two images made on my own, the one runs asterisk and the second one runs mysql with realtime asterisk config. After upgrade Docker from 1.13 to 1.13.1 in the evening i’ve got overfilled /var in the morning.
i’ve found huge container log filled with
{"log":"Asterisk Ready.\n","stream":"stdout","time":"2017-02-13T17:35:44.632544576Z"} {"log":"\u0000\u0000\u0000\u0000\u0000\u0000\u0000...\u0000","stream":"stdout","time":"2017-02-14T17:47:42.312998462Z"}
up to disk volume. records with \u0000 in them repeated
both images start with
docker run --net=host -d --name <name> --rm <image>
during further investigation i found that if to add -t to command, image runs normal without any problems. it is strange, i run images detached and -t is useless here.
who can help with this or explain? or it is an issue of version 1.13.1?


(Sebastiaan van Stijn) #2

Can you open an issue in the issue tracker? There are some differences in the way logging is handled with a tty attached, and without, but I’m not sure where these would come from. \0000 is unicode for NULL / NIL, but it’s a bit strange that’s being logged.

When opening the issue, please provide the output of docker version, and docker info, and it would be great if you could provide some information about the image you’re running (if public), and how (with what options) the container is started.

Note that for logging, it’s recommended to set a --log-opt max-size and --log-opt max-files so that log files are rotated (to prevent your host running out of disk space.


(Troff) #3

Hi!

Aready open https://github.com/docker/docker/issues/31308

Thank you, Peter