Docker-openldap: Startup fails after migration

Created on 4 Jul 2017 · 4Comments · Source: osixia/docker-openldap

I've just moved my ldap setup from one server to another by simply copying over all files (config, database in host directories, environment.yml and ssl certificates) and tried starting the container.
This fails with the following error:

Check previous TLS certificates...
Hi! I'm ssl-helper, what button should i press ?
cfssl-helper is launched, everybody on the floor!
Files /container/run/service/slapd/assets/certs/fullchain.pem and /container/run/service/slapd/assets/certs/key.pem exists, fix files permissions
Start OpenLDAP...
Waiting for OpenLDAP to start...
595be0a3 @(#) $OpenLDAP: slapd  (Jan 16 2016 23:00:08) $
    root@chimera:/tmp/buildd/openldap-2.4.40+dfsg/debian/build/servers/slapd
595be0a3 daemon: bind(7) failed errno=99 (Cannot assign requested address)
595be0a3 slapd stopped.

Any ideas what could be going wrong here?

(A simple docker run oisixia/openldap works by the way)

Source

jckoester

Most helpful comment

I had a similar issue solved by removing the domainname parameter from the docker-compose.yml file

jwpritchard on 22 Jun 2019

👍4 👎1

All 4 comments

After some more try and error, I believe this is a problem of docker-compose. The following composition worked well in coreOS but doesn't in debian stretch.

version: '2'

services:
    # dummy nginx for tls certificate generation
    ldap_nginx_dummy:
        container_name: ldap_nginx_dummy
        image: nginx
        restart: always
        environment:
            VIRTUAL_HOST: "ldap2.domain.tld"
            LETSENCRYPT_HOST: "ldap2.domain.tld"
            LETSENCRYPT_EMAIL: "[email protected]"
        volumes:
            - /data/ldap/nginx_conf.d:/etc/nginx/conf.d:ro
        networks:
            - outside
    # ldap container
    ldap:
        container_name: ldap
        image: osixia/openldap
        build:
            context: ./src/docker-openldap/image
        restart: always
        hostname: ldap2.domain.tld
        domainname: domain.tld
        volumes:
            - ./database:/var/lib/ldap
            - ./config:/etc/ldap/slapd.d
            - /data/ssl/ldap2.domain.tld:/container/service/slapd/assets/certs
            - ./environment:/container/environment/01-custom
        networks:
            - inside
            - outside
        ports:
            - "389:389"
            - "636:636"
        command: "--loglevel trace --copy-service"

networks:
    inside:
    outside:
        external:
            name: proxy-tier

In deboian this keeps bootlooping.
Running the docker command directly lwokrs perfectly well for me.

docker run --name ldap --hostname ldap2.domain.tld -p 389:389 -p 636:636 --net proxy-tier --volume /data/ldap/database:/var/lib/ldap --volume /data/ldap/config:/etc/ldap/slapd.d --volume /data/ssl/ldap2.domain.tld:/container/service/slapd/assets/certs --volume /data/ldap/environment:/container/environment/01-custom --rm dasmaeh/openldap --copy-service

I actually can't see the difference between the two.

jckoester on 9 Jul 2017

Unfortunately I met the same problem when i copied the config and database dir from a running docker-openldap node for replication. The replication operation is fine and runs well in test env but cannot start container in prod, it's so weird...