SRE Weekly Issue #315

In the previous articles in this series, they described a process of interviewing incident responders before a full retrospective meeting. This one discusses what to do if you can’t conduct those interviews, and the particular challenges this will bring and how to deal with them.

  Emily Ruppe — Jeli

Some interesting ideas on potential downsides of circuit breakers and how we might ameliorate them.

  Marc Brooker

GitHub has had a bit of a hard time lately. Here’s an update on what they’re dealing with and how they’re planning to address it.

  Keith Ballinger — GitHub

All sorts of “mean time to” metrics, including 6(!) different MTTR metrics and how they might be used.

  Alex Ewerlöf — InfoQ

This is a huge 100+-page report on the benefits of a model in which development teams own the operation of their systems. There’s a lot in here, with carefully spelled-out pros/cons and cost/benefit analyses. Need to convince someone? Send them this.

We’ve written this playbook for CxOs, product managers, delivery managers, and
operations managers.

  Bethan Timmins and Steve Smith — Equal Experts

It’s easy to miss MTUs, until they sneak up on you and cause really confusing problems.

  Aaron Kalair — Hudl

Should you compensate for on-call? How? I really want to see more articles about this, so send them my way if you see or write any.

  Chris Evans — Incident.io

Some good tips in this article, and I love the case studies.

  Prathamesh Sonpatki — Last9