A message from our sponsor, VictorOps:
Keeping your local repository in sync with an open-source GitHub repo can cause headaches. But, it can also lead to more flexible, resilient services. See how these techniques can help you maintain consistency between both environments:
http://try.victorops.com/sreweekly/keeping-github-and-local-repos-in-sync
Articles
This is a reply/follow-on/not-rebuttal to the article I linked to last week, Deploy on Fridays, or Don’t. I really love the vigorous discussion!
Charity Majors
And this is a reply to Charity’s earlier article, Friday Deploy Freezes Are Exactly Like Murdering Puppies. Keep it coming, folks!
Marko Bjelac
In this story from the archives, a well-meaning compiler optimizes away a NULL pointer check, yielding an exploitable kernel bug. I love complex systems (kinda).
Jonathan Corbet — LWN
A new report has been released about a major telecommunications outage last winter. This summary paints the picture of a classic complex systems failure.
Ronald Lewis
Making engineers responsible for their code and services in production offers multiple advantages—for the engineer as well as the code.
Julie Gunderson — PagerDuty
Outages
- Google Hangouts
- Google Cloud Platform: Followup for Incident #19020
- Incident #19020 occurred on October 22.
- Facebook, WhatsApp, and Instagram
- Yahoo Mail
- Second Life
