On-call Onboarding: The Ugly
ah, the final installment of on-call onboarding has arrived. previously I covered the good (pt1) and the bad (pt2) here's a brief sample of the ugly I've seen myself or heard anecdatally from folks
Finding out you're on-call (and what on-call is) by getting surprise paged into an active incident by your boss
- No runbooks || all runbooks are dead links
- No training or pairing with the observability tool(s)
- Owning way too many services || no ownership
- Inheriting services without a handover or documentation
- That orphaned service that no one really owns (aka by default ops will have to maintain it)
- Getting back to back to back paged while at a symphony and not being able to quiet my phone because I gave PD an override for silent mode and also was trying to not fall in my fancy dress as I stepped over tons of people as I dashed out #HugOpsToMe
- implicit 24/7 perma on-call
Pls share your tales of woe with me over on twitter or mastodon
Member discussion