
Kubeshark brings Kubernetes network traffic into incident response with eBPF capture, protocol-aware search, retrospective PCAPs, and MCP access for AI agents.
Blog about SRE, DevOps, Linux and Networks

Kubeshark brings Kubernetes network traffic into incident response with eBPF capture, protocol-aware search, retrospective PCAPs, and MCP access for AI agents.

Observability stack consolidation is becoming a practical SRE priority. The goal is not one magic tool, but one reliable incident timeline across logs, metrics, traces, alerts, and deploys.

OpenSRE gives SRE teams an open source framework for AI-assisted incident investigation. It connects alerts, telemetry, infrastructure context, runbooks, and chat workflows so responders can start with evidence instead of a bl...

Iroh gives platform teams a Rust networking stack for dialing machines by public key instead of chasing changing IP addresses. For SREs, that is useful wherever cloud, edge, mobile, and agent workloads need direct encrypted pa...

Large Postgres deletes can look like cleanup while quietly creating more work for readers, replicas, and autovacuum. For SRE teams managing retention, partitions and table drops are often the cleaner operational pattern.

Herdr treats AI agents like long-running terminal workloads: persistent, attachable, observable, and reachable over SSH. That is a practical direction for SRE teams running more than one agent at a time.