Skip to content

Common problems & FAQ

Bernhard Kaszt edited this page Jan 30, 2019 · 9 revisions

Common problems

The container "cms" keeps restarting

This can have multiple reasons:

The database doesn't exist

Make sure that db/docker-entrypoint-initdb.d was correctly mounted to /docker-entrypoint-initdb.d inside the db container.

You can check this by executing docker-compose exec db sh and stat /docker-entrypoint-initdb.

If this folder is empty, Docker failed to create the mount for it. Make sure that Docker can access your hard drive. If you are in a Windows domain, make sure that the password of your windows user has not expired. You might have to re-enter it.

Check the log file with docker-compose logs db. If it says "Installation of system tables failed!", add the line innodb_flush_method=O_DSYNC in db/conf.d/mysqld.cnf. Delete old files in db/data directory and run docker-compose up -d again (see https://github.com/docker-library/mariadb/issues/38).

The database structure can't be created

If the database was deleted, make sure to delete all files in cms/dbfiles/, because the database import will fail when this folder is not empty.

The CMS can't be accessed in my browser

Sometimes, the Apache process fails to restart in the cms container. At the moment we do not know what causes this. It seems to be Docker bug.

You can check this with docker-compose exec cms bash and then supervisorctl status.

If service apache2 is not running and service tomcat keeps restarting, you can probably fix this with supervisorctl start apache2.

Full hard disk

Containers fail to start when your hard disk is full. Be sure that you have sufficient disk space.

The container "elastic" doesn't start

Be sure that you have enough memory (default setting is 2GB) available on your system. When using Virtualbox, it might be necessary to give the Virtual box container more memory.

max virtual memory areas vm.max_map_count [65530] is too low

See: https://www.elastic.co/guide/en/elasticsearch/reference/current/docker.html#docker-cli-run-prod-mode

command xyz returned a non-zero code:127

This error means that a command (or shebang interpreter in a shell script) was not found.

Most times this is caused by cloning the GIT repository with windows line endings. This breaks the linux shell scripts. However this can be fixed easily by running git config --global core.autocrlf input before cloning the GIT repository.

If it's already broken, you can also try to convert the windows line endings of all *.sh files. You maybe have to rebuild the docker images with docker-compose build --no-cache afterwards.

FAQ

How do I reset the CMS?

  • docker-compose down
  • Delete all files in cms/dbfiles/
  • Delete all files in db/data/