Reboot of working docker and awx stops containers and cannot start or re-install

Archives AWX Project
,
1

Hello All,

This is getting old. the base problem that I was having was the awx_postgres container would
go into a restarting loop - I could not fix it.

I attempted to modify the inventory file to give a location out side of /root/.awx/pgdocker (and awxcompose)

I tried to run the install playbook again - I am not connected to the internet. the install errors out saying it is a “read only filesystem”

I then uninstalled the docker packages and then reinstalled following this link: https:docs dot docker dot com/engine/install/ubuntu

I was able to reinstall docker, docker-ce, etc all the apps according to the link above. I was also able to do the docker run hello-world program
It returned a success message (it must have made it thru the proxy)

I am using Ubuntu 19.10

Then I went to the installed dir for awx and ran the installer - it worked - took a while - awx worked fine - then I shut machine down to take
a snapshot - rebooted - and the entier install went away - none of the containers were started. I attempted to run the install.yml program again using the

folders outsite /root/.awx (which worked before the reboot) and get a message now "cannot install, etc /AWX/pgdocker a read-only file system.

What does it take to get this application to work?

any suggestions? I have also the the awx_postgres stuck in the “restarting” loop.

thanks

ewholz

2

Eric,

If you are willing to reinstall, have you tried the official ansible
role from AWX to do the installation?
https://github.com/ansible/awx/blob/devel/INSTALL.md

3

Well you have to update packages or reboot at some point, and you
won't know if it's broken until you try. Better now than after a bunch
of work creating users, templates, credentials, etc. And you would
have the docker-compose file, and all the credentials from
credentials.py, environment.sh, SECRET_KEY, etc. for portability if
you had to move to a different machine.

I would make sure to disable updates of docker and its dependencies at
least for the first time around patching/rebooting. It could be a
newer (or older) version of docker or docker-compose or containerd or
some other package causing issues like with which Storage Drive docker
uses. Then if you reboot and all is fine then great. If not, then
look at what packages were updated and debug from there (roll back,
etc.).

4

Hello Mgztls,
Thanks for the reply - this is a real problem with docker - I did a rebuild 5 days ago, and all was solid.
Then today - the containers were no longer running - and of course, trying to run intall.yml failed as the directories for pg_docker and awxcompose
were already there. It seems the install file always wants to create new directories for those items. Is there some way to get the intall program

to work without trying to recreate those directories??

Thanks for the info,

ewholz

5

Eric,

After deploying the awx stack via the instructions @ https://github.com/ansible/awx/blob/devel/INSTALL.md how are you starting/stopping the containers thereafter? It sounds like you are trying to run install.yml every time you just want to start, stop, or restart the awx stack. You know that after it’s been built and running that you would just use ‘docker-compose’ from then on to manage the awx stack, right?