Startup time can be really abysmal: Almost 600s for loading the image
(on a fresh node) and some time to run the initial setup.

Address #60

We noticed the subprocess that runs our 'cron' sometimes quits silently;
one possibility we see is that OOM kills it. In any case, we now also
check for the existence of the commands `sleep` or `timeout` in the
process table (`ps`) via the `livenessProbe`.

Address #60

Lately, we encountered hick-ups when the dspace container comes up but
the postgres container is still missing (although the containers will
restart until both start in the right order).

Here we add one `initContainers` that will poll the postgres db until
success.

Address #60

'502 Bad Gateway' errors from the k8s ingress server can be solved
(sometimes) by restarting the nginx pod. Here we check
https://clarin(-dev).eurac.edu:443

Addresses #53

- disabled ssl-passthrough
- configured ssl-certs

FIXME:
- repository/oai is not accessible via http
- battle test the config

fixed.

- ingress also listens on port 80
- nginx redirects to https except for path: repository/oai

- bumped version numbers to 1.0

- bumped version number in kube deploy

  version to "latest"

- removed superfluous yaml files