What does a real rollback plan for a model require?
mediumAnswer
- The previous model artefact still deployable, which means versioned artefacts with their exact dependency set, not just a file in object storage.
- It also requires the previous feature pipeline, because a rollback that feeds the old model new-format features is not a rollback.
- A configuration switch that routes traffic without a code deployment, so recovery takes seconds rather than a release cycle.
- A defined trigger, stated in advance as a metric and threshold, since deciding whether to roll back during an incident wastes the time you do not have.
- And an explicit answer for state already written by the bad model, such as decisions applied to accounts, because reverting the model does not revert its effects.
- Rehearse it, or it does not exist.
Check yourself — multiple choice
- Keeping the old model file somewhere
- A versioned artefact with its dependencies, the matching feature pipeline, a config-only traffic switch, a pre-agreed trigger, a plan for state the bad model wrote, and a rehearsal
- A git revert
- Rollbacks are unnecessary with canaries
Rollback needs the artefact, the matching pipeline, a fast switch, a pre-agreed trigger and a plan for side effects.
#deployment#reproducibility
Practise MLOps & Data Quality
215 interview questions in this topic.