A little bit of everything this week, with an emphasis on logging, Kubernetes, and time-series databases. Enjoy! 🍩☕📈
This issue is sponsored by:
What does the future of OpenTelemetry hold? Chronosphere's Co-founder and CEO, Martin Mao, breaks down his predictions for open standards, open protocols and more specifically, OpenTelemetry in our newest video series, Observability for What’s Next. Watch the clip here.
Articles & News on monitoring.love
Come hang out with all your fellow Monitoring Weekly readers. I mean, I’m also there, but I’m sure everyone else is way cooler.
From The Community
Because everyone knows that having six pillars is twice as good as only having three. I bet I could come up with another few and venture capitalists would be banging my door down.
P.S. All kidding aside, this is a great review of the “pillars of observability”, adding more context and broadening our definition of the data we use to understand our systems better.
We read about time-series databases all the time, with an implicit acknowledgement that they’re different from traditional relational databases. Here’s a helpful post that compares their raisons d’être.
An excellent article on Kubernetes Pods eviction, along with a pointer on how to monitor them in Prometheus.
Some tips and best practices to consider if you’re able to log in JSON (aka events).
Good culture should always strive for a sustainable environment for employees, but it’s easy to take traditional responsibilities (e.g. on-call) for granted. This article is a good reminder of the behaviors we want to model and incentivize among our engineering teams.
There are some really insightful bits in here, particularly about how we reason about complex systems (with a nod to Dr. Richard Cook), before veering off the rails into a predictable “existing tools suck, buy mine” spiel (sigh).
As an ex-Herokai and long-time Heroku fan, I’m excited to see support added to Grafana Loki to accept log drains from the popular PaaS offering.
If you’re looking to run a pure open source monitoring stack, this guide should help you get started. No big surprises here, but it covers basically all the tools you’d expect.
Moderate severity fixes affecting Grafana Auth Proxy and RBAC. There appears to be a manual workaround to mitigate the RBAC privilege escalation attack vector.
If you work with any teams using Apache Spark on Amazon EMR, here’s a helpful pattern for monitoring those Spark jobs.
See you next week!
– Jason (@obfuscurity) Monitoring Weekly Editor