seqr deployment fails; timeouts. Centos 7.8 #1409
Description
Greetings. With a file docker-compose.yml in place, and python 3.8.1 / docker compose in path, I follow these steps:
sudo docker-compose up -d seqr
sudo /usr/local/docker-compose/docker-compose up -d seqr
However, reviewing those logs confirms that the first step never actually finishes, despite the promising output below.
Creating network "app_datasets_default" with the default driver
Creating app_datasets_redis_1 ... done
Creating app_datasets_elasticsearch_1 ... done
Creating app_datasets_postgres_1 ... done
Creating app_datasets_seqr_1 ... done
The logs show that all net-related download actions within the docker are timing out. Example:
WARNING: Retrying (Retry(total=4, connect=None, read=None, redirect=None, status=None)) after connection broken by 'ConnectTimeoutError(<pip._vendor.urllib3.connection.VerifiedHTTPSConnection object at 0x7f0ab3876450>, 'Connection to pypi.org timed out. (connect timeout=15)')': /simple/django/
Adding proxy info to a config ( /root/.docker/config.json, reference: https://medium.com/@airman604/getting-docker-to-work-with-a-proxy-server-fadec841194e) will not avail. Everything halts with thingvs like"Container "4e19a6c5f903" is unhealthy."
In addition to the timeouts, I see mention of a non-existent host postgres. I assume that was to have been created within the docker environment:
psycopg2.OperationalError: could not translate host name "postgres" to address: Name or service not known
The environment is an oVirt VM running Centos 7.8.
Do you have any leads on the conditions needed to get seqr running? We there special firewall exceptions needed outside of docker?