Ark Protocol·10h agoOur Outage Was Not a Bug. Our Kubernetes Liveness Probe Killed 5 Healthy PodsThere was no bug in our outage.
Ednilson Correa·12h agoFilelog em protobuf: medi o ganho real e achei uma armadilha no rollbackO benchmark promete 5x. Em um Collector inteiro, foram 21% de CPU, e voltar para uma versão antiga pode duplicar seus logs.
Seungmin Lee·21h agoI Built It to Survive Flash Sales. Now Customers Pay for It.How my SRE automation for flash sales became a feature customers book themselves
Adam Jaber·22h agoYour Health Check Killed the Servers It Was CheckingA /health endpoint that checks the database feels responsible. Then the database blips, every instance fails the same check in the same…A response icon1A response icon1
InITNEXTbyomrilotan·1d agoWho’s Watching Production When the Agents Ship?Site reliability lessons for a one-person company of agents
InUnicobyLeandro Ikeda Santos·1d agoFrom SLO to SLA: when “all green” isn’t enoughHow Unico started calculating availability the way customers experience it
Noah Grinstein·1d agoHow to Test Failover Without Taking Production Down: Game Days, Chaos Experiments, and a…Test failover safely by climbing a ladder from tabletop exercises and staging tests to small, reversible production experiments with a…
Gabriel Okiri·1d agoThe Deploy That Went Green and Changed NothingThe pipeline said the release had shipped. The cluster was still running last week’s image.
Gabriel Okiri·1d agoThe App Was Talking. Nobody Was Listening.The pod was healthy, the logs were live, and the dashboard was empty for three days.
Tinguyen·1d agoYour App Is Up. Can You Still Ship a Fix?A healthy production dashboard can hide a broken path from finding a bug to safely fixing it.
