Stack Overflow SRE with Mark Henderson

Episode #1 | Published August 22, 2018

Mark Henderson joins us from the Stack Overflow SRE team.

Show Notes

Mark Henderson joins us from the Stack Overflow SRE team to discuss a bit about the SRE team, what Stack Overflow's SREs are responsible, and how Stack Overflow uses Elastic.

The SRE team ingests between 600-800GB per day to their logging cluster and uses that data for triage and troubleshooting.

Stack Overflow manages their instances using puppet and have an automated patching system used for automated system patching.

You can find Mark Henderson on Twitter at @thefarseeker.


Question of the Week

Why does Elastic recommend not spanning multiple datacenters within a single Elasticsearch cluster?

Being a distributed system, Elasticsearch needs reliable, low-latency connections between nodes within a cluster. Leaving the datacenter tends to hurt both of the options.

You can read more here.