On-call

Leigh asked:

Anyone read or have any good advice for organising out of hours cover for small dev teams managing their own services? E.G. how to fairly compensate for overtime, organise rotas, manage escalation of issues, etc?

What we used at Skype seemed to work well.

The team was split between London and the US West Coast. That covered most of the day. Only the middle of the night (UK time) needed to be covered.

We spent one week on call. The team was eight people, so that happened about every two months. This is where having a larger team helped us.

Each week had a primary and secondary person. The following week, the secondary would become the primary, and the next person on the rota would be the secondary. The primary got ~£500 plus time off in lieu for any public holidays during the week. The primary got the payment and time off regardless. The secondary only got the money and time off if they were called out.

Our SLAs mean the on-call person could still go to the cinema. As long as they were at most a couple of hours from their laptop

The Increment has some decent recommendations.

  • Category: SRE