# RIOT CI meeting 6/22 ## Details When: 20.06.22 10h Berlin time Where: https://meet.jit.si/RealFlowersLaughFlatly Previous meeting notes: https://hackmd.io/GOvgvGGeTkCpP7hwiGFImg ## Participants - Kevin - Kaspar ## Agenda - Migration status update - Deploying spare Raspi 4s for the pi fleet - How to switch between staging & prod (or revert to old murdock) ## Notes ### Migration status update - ci-prod, ci-staging in place, working fine - TODO: - [x] https://github.com/RIOT-OS/RIOT/issues/18103 (not murdock-ng specific, but single worker trips on it) - [x] no pifleet -> RUN_TESTS=1 builds stalling - [x] Get ci-prod reliably running a subset on 1 worker (left: fix #18103) - [ ] Enable on Github status - [ ] Run board subset on ci-prod matching all pi fleet and remove pi fleet from old murdock (deploy with either salt or ansible) - [ ] => ci-prod is now handling the pifleet-based HIL tests - [ ] External people contribute to pi-fleet - aabadie fill fix a small bug in the cleanup API (which is why there are so many memory full warnings) - Kaspar will provide a nice CLI for murdock management (written in rust) ### Deploying spare Raspi 4s for the pi fleet An image that works on all the pis (still needs UDP for networking). TODO: KW to ask HAW IT to allow UDP (fixed host with fixed ip). Currently openVPN migrating to wireguard. [pi fleet](https://github.com/kaspar030/pifleet-minion) Assuming Raspi 4s work with 32bit image, let's go with those first, then update the pifleet-minion scripts and salt config (some things assume 32bit arm). ### How to switch workers between prod, staging, old murdock ssh config just needs to be changed -> hostname, keys. Just FYI -> kaspar will be out for 2 months living life (but also with his laptop).