On April 8 – 13, I attended the Infrastructure Hackfest (that everyone keeps calling “Hackathon“, but in my mind it’s always been a hackfest).
The hackfest was in Paul Frields (stickster)’s hometown of Fredericksburg, Virginia (as opposed to Frederick, Maryland).
As usual, we had a published list of goals on the Fedora Wiki.
Also, as usual, the list of goals got modified somewhat during the hackfest: the ELK goal got scrapped, and I was roped into some ostree/atomic goals.
On Monday, we spent a good while discussing the new workflow for Fedora packagers, which has changed a lot the last year due to Arbitrary Branching, Pagure on dist-git and PDC (which we were informed was going to be deprecated, fun) and how to improve the workflows and their documentation. I also worked with Randy to get a token for WaiverDB for Bodhi (after which we found out that the code didn’t send the right API calls, but we will get that fixed soon).
On Tuesday, we had some discussions over application legacy of a team member who switched teams, and retiring of some older applications (goodbye, summershum and darkserver).
We also discussed some things around Rawhide gating, and got a full idea on both the high-level overview and how implementation in Bodhi would come to be.
On Wednesday, we started setting up AWX (upstream Ansible Tower), but this is also where I was roped into various ostree things (Atomic Host artifacts during Bodhi composes and ostree mirroring via CloudFront).
We hit some snags with the AWX setup, and we’ll be discussing this further in the infra team soon.
On Thursday, there were more discussions about rawhide gating, and I gave a quick overview of some important OpenShift concepts that are useful for our app developers. I’ve been told that quick tutorial was very useful, and we decided to try it out on the spot: we got most of Bodhi’s web UI’s move to openshift prepared.
On Friday, everyone was traveling, so those of us leaving later in the day met up in the hotel lobby.
Me and Randy actually got Bodhi’s web UI fully working in Openshift (both staging and prod!) before he had to leave.
We switched the Staging instance fully over, but Production we left for another day (since a prod deployment on Friday is not a good idea…).
All in all, I had a lot of fun and think we made some good progress on most of the set goals (and also some non-pre-set goals).
I’d like to thank Paul for pulling this together, and for being a good host in his hometown!