Open velle opened 1 week ago
When I first posted this issue, I actually did not try out if the services were running or working at all, I just noted the "FATAL" error and assumed it was not running properly . Today I tried out if it actually worked.
All four containers are up and running, and it's possible to connect to all of them with tty.
I'm able to open http://localhost:8080/dashboard/sign-in, but if I sign in with user gc2 and password 1234, I get the following error.
Even after pruning everything in docker-compose, the first time I run docker-compose up
, the postgis container (gc2-vidi-docker-compose-postgis-1) prints:
****************************************************
INFO: GC2 system already initiated.
Doing nothing else than start the service.
****************************************************
I believe this is printed from https://github.com/mapcentia/geocloud2/blob/87a6112f6a6d961467f34b1e060800a443193569/docker/postgis/entrypoint.sh#L13.
Am I on to the right cause of the problem? I was expecting it to fail that if-test and then do all the "First run" stuff instead.
I have had issues stopping some of the containers, as shown in the first posts screenshot. They simply would not stop, no matter what I tried, and I got "Error while stopping" and "Error while killing".
I found out this was caused by AppArmor, and I fixed it with:
sudo dpkg-reconfigure apparmor
sudo apparmor_parser -r /var/lib/snapd/apparmor/profiles/snap.docker.compose
Now all problems with stopping containers are gone. However, FATAL: role "root" does not exist
persists.
I cloned this repo and fetched the images a few days ago, and ran into this problem. The containers keep trying stuff and print status/errors forever.
To reproduce:
Complete log from docker-compose: https://gist.github.com/velle/cfdc9c7c5039f2468f044a3dd89c709a#file-gistfile1-txt
Details: