Thread by @lizthegrey

About to go onstage in the Gold Room at the Palace Hotel to tell the world how Stackdriver and GCP make implementing SRE practices easier so you can have high reliability! #GoogleNext18

Woo, we did it! And I made it to my next session on time afterwards too!

Summary: we discussed four key areas: (1) why operability matters and how DevOps+SRE improve reliability, (2) why/how to set up SLOs & an error budget focused on measuring your customers' needs, (3) reducing TTD/R w/ tooling, and (4) transparent SLIs for debugging #GoogleNext18

Key tech: Stackdriver SLO for instrumenting, measuring, and alerting on error budgets. Stackdriver Monitoring features including Metrics Explorer by locality tags, metadata joins, and Query Notation for advanced joins. Exemplars for traces and metrics in one view. #GoogleNext18

All of these inspired by the best practices of 15+ years of SRE at Google. We want to make it easier for everyone to scale their systems without overloading their ops teams. #GoogleNext18

and we want to improve collaboration with our customers by reducing silos not just within companies but between customers and their cloud providers. Thus, transparent SLIs from us to you, and, soon, transparent SLIs from you to us if you want a single pane of glass. #GoogleNext18

Here are the slides for the talk on implementing SRE with help from Stackdriver tooling!

I'll link the YouTube video as soon as it's available as well. [btw this thread is one of the few times on Twitter that I *do* speak for my employer] drive.google.com/file/d/1iOMaYI… #GoogleNext18

As much as SRE is about doing more with tooling, it's also about culture change! Remember to check out cloud.withgoogle.com/next18/sf/sess… tomorrow about the cultural changes needed to make SRE work! #GoogleNext18

Session video is up for how to improve reliability by implementing SRE measurement and debugging techniques with Stackdriver!

Kudos to the video team for producing it on such a short turnaround! #GoogleNEXT18