Open mark-barnes opened 4 years ago
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
Reopening, we probably should have disabled stale bot while we were on hiatus.
Do we need to mention creating CRs (or whatever mechanism is prefered) and making sure they are peer reviewed and change is communicated etc? Engineering principles of testing/backing up/communicating change and having a rollback plan? Is this two competencies ?
there are some examples recently (https://medium.com/ft-product-technology/a-chaos-test-too-far-6405c2cc4bbb) and the dns incident where having a known recent good position to go back to easily would have been pretty handy