make `fastcgi_read_timeout` in nginx configurable to increase stream connection while import POST requests
Signed-off-by: Frank Kloeker <frank.kloeker@telekom.de.de>
- Move the healthcheck from the docker-compose example to the actual image. That makes sure all user of the image automatically get the healthcheck.
- Change to `curl --fail || exit 1` as suggested in [docker documentation](https://docs.docker.com/reference/dockerfile/#healthcheck)
- Add `--silent --show-error` so that docker health state contains the output of the HTTP call instead of curl progress bar
- Set the user agent to make the logs more readable
Most example docker-compose files include restart: unless-stopped or restart: always. This being absent on Wallabag's example leads to unexpected downtime upon server restart
This PR adds in a health check that can used for reporting, monitoring, or autohealing, as well as a `depends_on` statement that ensures wallabag comes up only once the DB and Redis are up.
Update timeout/retries to be more sane
The default value of this variable leads to a lot of issues with assets
when new comers try wallabag. This change should reduce the occurrence of
this kind of issue.
Signed-off-by: Kevin Decherf <kevin@kdecherf.com>
* MariaDB encoding is set when creating DB through the entrypoint
* Cleaning Dockerfile
* Removed `POSTGRES_USER`
* Combined run commands
* Changed deprecated `MAINTAINER` to `LABEL`
somehow there is a problem with the parameters.yml and environment
variables. ansible templates are now used.
database migration is still a big problem.
