1 overview.tux
2 alerts.log
3 pods.yaml
buffers: 3
┤ latency p99 (ms) — OK ├
10s agonow 42ms
┤ cluster ├
availability
100.000%
pods
██████████ 42/42
pages resolved
0
coffee
██████████ 100%
┤ journal ├
┤ pods.yaml — READ-ONLY ├
apiVersion: v1 kind: PodList metadata: # resourceVersion is a lie kubectl tells itself resourceVersion: "1337" items: - metadata: name: tux-prod-7f9 namespace: prod labels: app: tux mood: fine-probably spec: containers: - name: tux image: tux:v1.4.2 # DO NOT roll back. (we always roll back.) resources: requests: coffee: 100m patience: 0 status: phase: Running # ^ aspirational. see restartCount. restartCount: 137 # same number as the exit code. no, not a coincidence. lastState: terminated: reason: OOMKilled # it dreamed too big exitCode: 137 - metadata: name: tux-prod-2b1 namespace: prod status: phase: Running restartCount: 0 note: "the reliable one. nobody thanks it." - metadata: name: tux-canary-9k2 namespace: prod annotations: deployed-by: "intern" deployed-on: "friday" # brave. foolish. brave. status: phase: CrashLoopBackOff restartCount: 9001 conditions: - type: Ready status: "False" # narrator: it was not ready - metadata: name: coffee-machine-sidecar namespace: prod # not in the architecture diagram. load-bearing anyway. status: phase: Running restartCount: 0 criticality: maximum - metadata: name: dns-resolver-4x1 namespace: prod status: phase: Running # it's fine. it's always fine. until it's DNS. restartCount: 3 # kubectl get pods -o yaml # do not edit. regenerates itself from the cluster's trauma every 30s.
NORMAL
prod-cluster.yaml
utf-8
yaml
00:00
:checkhealth prod