- 보내짐
- 2026년 11월 17일 오전 2:15 UTC
- 작성자
- Kamo
- 커밋
- 0a83a96
maxSurge 1 / maxUnavailable 0 with the readiness probe already start the new pod and wait for it to be ready before the old one stops, so a deploy never drops the service. Outside a deploy one pod is enough; the second replica only took memory, and k1m1 running out of page cache was what kept stalling etcd and the control plane (2026-11-17).
