Provide and deploy systemd service files so that services restart when the server(s) reboots

[candleindark]

We currently have a cron job available to restart the full feature instance at typhon using the script at https://github.com/datalad/datalad-registry/blob/master/tools/service.sh. We don't currently have that on falkor for the public instance.

[yarikoptic]

don't do cron job -- make it into a "proper" systemd service. E.g. here is how it looks for reprostim

check e.g. on typhon service files and README under /data/repronim/reprostim-reproiner/code/services for reprostim.

edit: yes, do for both typhon and falkor (web read-only). Commit those service files in here.