When adding support for new models, or naming conventions that might not pattern-match the ones we have included it would require editing the config.py list.
Describe the solution you'd like
It's OK to have a baseline set of systems kept in config.py probably, but this should also be user-configurable to include additional models or types in /opt/quads/conf/quads.yml
Then, when quads-server is started and quads.yml is parsed into memory it should append the existing lists of SUPPORTED and SUPERMICRO respectively with the example user input below.
Placing duplicate model strings should pick the first one and not cause an error if it already exists in config.py already.
Is your feature request related to a problem? Please describe.
We currently keep a list of badfish-capable models (Dells) here:
https://github.com/redhat-performance/quads/blob/master/quads/config.py#L29
We also keep a list of supermicro types (to use ipmitool instead of badfish as you need a paid license to use badfish API on SuperMicro).
https://github.com/redhat-performance/quads/blob/master/quads/config.py#L30
When adding support for new models, or naming conventions that might not pattern-match the ones we have included it would require editing the
config.py
list.Describe the solution you'd like
It's OK to have a baseline set of systems kept in
config.py
probably, but this should also be user-configurable to include additional models or types in/opt/quads/conf/quads.yml
Then, when
quads-server
is started andquads.yml
is parsed into memory it should append the existing lists ofSUPPORTED
andSUPERMICRO
respectively with the example user input below.Placing duplicate model strings should pick the first one and not cause an error if it already exists in
config.py
already.e.g.