Merge lp://qastaging/~dpigott/lava-lab/make-lava-lab-have-all-device-configs into lp://qastaging/lava-lab
Status: | Needs review |
---|---|
Proposed branch: | lp://qastaging/~dpigott/lava-lab/make-lava-lab-have-all-device-configs |
Merge into: | lp://qastaging/lava-lab |
Diff against target: |
815 lines (+725/-0) 18 files modified
lava/device-types/beagle-xm.conf (+23/-0) lava/device-types/beaglebone-black.conf (+21/-0) lava/device-types/beaglebone.conf (+21/-0) lava/device-types/capri.conf (+46/-0) lava/device-types/highbank.conf (+2/-0) lava/device-types/k3v2.conf (+46/-0) lava/device-types/kvm.conf (+3/-0) lava/device-types/mx51evk.conf (+11/-0) lava/device-types/mx53loco.conf (+30/-0) lava/device-types/nexus.conf (+46/-0) lava/device-types/nexus10.conf (+46/-0) lava/device-types/qemu.conf (+7/-0) lava/device-types/rtsm_foundation-armv8.conf (+20/-0) lava/device-types/rtsm_ve-a15x1-a7x1.conf (+111/-0) lava/device-types/rtsm_ve-a15x4-a7x4.conf (+111/-0) lava/device-types/rtsm_ve-armv8.conf (+128/-0) lava/device-types/vexpress-tc2.conf (+36/-0) lava/device-types/vexpress.conf (+17/-0) |
To merge this branch: | bzr merge lp://qastaging/~dpigott/lava-lab/make-lava-lab-have-all-device-configs |
Related bugs: |
Reviewer | Review Type | Date Requested | Status |
---|---|---|---|
Linaro Validation Team | Pending | ||
Review via email:
|
Description of the change
ok - so we have device-type configs in two places - the dispatcher and lava-lab, and any values in lava-lab override ones in dispatcher. What I've done is to make sure that lava-lab is up to date, taking care not to override any existing configs.
It strikes me that we have another problem to solve: Having device configs in two places, which are out of step. lava-lab needs to hold the current set, and any changes to it need to be reflected back into the dispatcher. The reason we need lava-lab to be definitive is that means we can re-salt to update a device type, rather than having to deploy a potentially unstable dispatcher.
I'm not self approving this, because it needs discussion, but it's already resulted in a fastmodels device failing because its dispatcher was wound back too far and the dispatcher didn't have the config for it and neither did lava-lab.
Unmerged revisions
- 200. By Dave Pigott
-
Update to latest
- 199. By Dave Pigott
-
Make lava-lab the definitive device-type list