I have 11 Docker containers that are set to autostart, but after upgrading to beta14b the docker won't autostart after a reboot.

Turning Docker off and then on again starts all the containers.

This last time I rebooted, one of the containers (Plex Media Server) got started automatically, but the rest didn't.

In the syslog there are a couple of lines pointing to an error, that lookes like this:

 

Mar 10 16:16:48 Tower logger: el="fatal" msg="Error: failed to start one or more containers"MariaDB: Error response from daemon: Cannot start container MariaDB: Error starting userland proxy: listen tcp 192.168.1.15:3306: bind: cannot assign requested address time="2015-03-10T16:16:48+01:00" level="fatal" msg="Error: failed to start one or more containers"nsenter: started succesfully!

 

Is there a timing/startup order issue, something to do with dependencies?

syslog.txt

Come to think of it, the Plex Media Center container is the only one that's got the "privileged" setting enabled. Maybe that helps narrow it down...

I have 11 Docker containers that are set to autostart, but after upgrading to beta14b the docker won't autostart after a reboot.

Turning Docker off and then on again starts all the containers.

This last time I rebooted, one of the containers (Plex Media Server) got started automatically, but the rest didn't.

In the syslog there are a couple of lines pointing to an error, that lookes like this:

 

Mar 10 16:16:48 Tower logger: el="fatal" msg="Error: failed to start one or more containers"MariaDB: Error response from daemon: Cannot start container MariaDB: Error starting userland proxy: listen tcp 192.168.1.15:3306: bind: cannot assign requested address time="2015-03-10T16:16:48+01:00" level="fatal" msg="Error: failed to start one or more containers"nsenter: started succesfully!

 

Is there a timing/startup order issue, something to do with dependencies?

 

Can you please remove all plugins from your system, reboot, and see if the error occurs?  You can leave containers / VMs alone, but just remove all the extra plugins you have installed (cache dirs, powerdown, nerdpack, etc.)

Same result, only the Plex Media Server (PMS) container is started.

Of the 10 containers that are set to autostart, only Plex Media Server has the setting "Privileged" enabled.

After disabling/enabling Docker, all the containers are started without errors.

syslog.txt

Same result, only the Plex Media Server (PMS) container is started.

Of the 10 containers that are set to autostart, only Plex Media Server has the setting "Privileged" enabled.

After disabling/enabling Docker, all the containers are started without errors.

 

Can you try disabling autostart on your containers and enabling them one by one to help track down which one may be causing the hang up?  Sounds like something that is specific to an individual container you have loaded.  Would be helpful to know which one so I can try to recreate this on a test system.

I disabled autostart on all containers, and then enabled only one container at a time and rebooted (10 reboots).

The only time a container was autostarted, was when Plex Media Server was selected to autostart.

I've attached two syslogs, the one with Plex (autostart worked) and the one with Transmission (autostart failed).

The syslog of the other boot cycles all had the following error, with only the names of the containers different:

 

Tower logger: Transmission: Error response from daemon: Cannot start container Transmission: Error starting userland proxy: listen tcp 192.168.1.15:54321: bind: cannot assign requested address

syslog-Plex.txt

syslog-Transmission.txt

I disabled autostart on all containers, and then enabled only one container at a time and rebooted (10 reboots).

The only time a container was autostarted, was when Plex Media Server was selected to autostart.

I've attached two syslogs, the one with Plex (autostart worked) and the one with Transmission (autostart failed).

The syslog of the other boot cycles all had the following error, with only the names of the containers different:

 

Tower logger: Transmission: Error response from daemon: Cannot start container Transmission: Error starting userland proxy: listen tcp 192.168.1.15:54321: bind: cannot assign requested address

Thank you so much for the testing. I will attempt to recreate tomorrow.

Did you re-create you Docker image and re-add you user defined templates. It'd just take a little time to re-create it all and see if that gets rid of the issue.

 

http://lime-technology.com/forum/index.php?topic=38323.0