Operations | Monitoring | ITSM | DevOps | Cloud

ClickHouse monitoring guide: the Prometheus endpoint, the system tables that explain slow queries, the Too many parts limits, replication lag and alerts. Sejal Pandey works on content and growth at Last9, writing about observability, reliability, and SRE practices.

GKE monitoring guide: what Google Kubernetes Engine collects by default, which metric packages to turn on, Autopilot limits, and the alerts to set. Sejal Pandey works on content and growth at Last9, writing about observability, reliability, and SRE practices.

✓ operational truth IT finds out late because traditional monitoring is built to detect what goes wrong, not what has quietly stopped happening. Closing that gap requires observability that validates business journeys end to end, detects missing activity, checks its own coverage, and predicts degradation before a threshold is ever crossed.

Most teams that run on Azure do not run only on Azure. There is a database on a VM nobody wants to move, a Kubernetes cluster somewhere else, a few servers in a rack, and a monitoring setup that grew around all of it. Azure Monitor sees the Azure part very well and nothing else, so the picture of an incident ends up split across two consoles, two alerting configurations and two sets of dashboards. Bleemeo now connects to Azure the same way it already connects to AWS.

No one can guarantee that a website or service never goes down. What you can promise is a measured availability target, and with a multi-location, active-active design you can reach 99.999% (five nines), about 5 minutes 15 seconds of downtime a year. That takes redundancy at every layer, automatic health-based failover across regions and ideally providers, safe deployments, failure testing, and outside-in monitoring. Control Plane is built for that tier.

At incident.io we are huge fans of Postgres; we've written about it a lot over the years, including how to choose the right indexes and how we're proud of being boring (The Pet Shop Boys). We use Postgres as our primary transactional database, which as of today has ~900 tables, and counting! The vast majority of our codebase does something along the following lines: read some data from Postgres, execute some business logic, then write that data back to Postgres. It is not, however, always that simple.

AI-assisted development can increase the rate at which teams produce code, but teams only realize those velocity gains if CI can keep pace. More pull requests (PRs) mean more builds, tests, and pipeline executions. Slow jobs leave developers and coding agents waiting for feedback, flaky failures consume time in reruns and investigations, and unnecessary test execution increases runner demand as delivery volume grows.