Those are great ideas. In time I think we can do all of them. My plan was to start with python imports as a way to maintain the hh5 environment.
As for transition, I think the stuff on the postgres database would be fairly straightforward to do without any downtime. Its a series of ‘CREATE ROLE …’ and ‘ALTER TABLE …’ postgres commands. Then I’d need to dig into the ansible config for jenkins7 and the puppet config for accessdev to enact the firewall/reverse proxy changes. Stats gathering can run as-is in the background while all of this is happening. Once that’s all complete, I can get rid of start_tunnel in dumpstats, and modify the config file to point to the new database URL (same database, just a different URL to get to it). After that, start working on the updated dump_* tasks.