Files
cluster-map/changelog.yaml
T
Виталий ЧефановandClaude Opus 5.5 3cdc2583ee v0.8.0: problem badges, actions as the viewer, load, events, lock
Problems: cards, namespace and node boxes and the tree carry a yellow or
red badge, computed on the server from object state (crash loops, image
pulls, pending and unschedulable pods, readiness older than 2 minutes, a
stuck rollout, a failed CronJob run, Services without endpoints,
Ingresses to missing Services, node conditions), recent Warning events
(notable reasons at once, others after five repeats) and Deckhouse
ClusterAlerts mapped by their labels. A Problems panel lists them all.
Checked against the live cluster: upmeter's synthetic probe pods and
ingress-nginx's fakeservice Ingresses are excluded, they fail on purpose.

Actions: restart, scale, run a CronJob now, suspend/resume, delete a
pod, cordon/uncordon. They run with the viewer's bearer token from the
DexAuthenticator (sendAuthorizationHeader), buttons follow a
SelfSubjectAccessReview, the service account stays read-only. Cross-site
writes are refused by Sec-Fetch-Site/Origin/content type. Scale retries
on conflict: right after a restart a single Get+Update got 409.

Load (metrics.k8s.io against requests/limits/allocatable, per-pod table)
and an Events tab. Cards are locked until the padlock is opened; the map
exports to PNG; the page refreshes badges every minute in place.

Performance, measured on the live cluster:
- layout of big views 3-4x faster: equal-size cards with no edges are
  laid out on the grid ELK would produce (same sizes, checked against
  ELK on a frozen node view), skipping ~8 ms per ELK call;
- hubble-relay gets ~20% fewer flows: the to-overlay copy of pod-to-pod
  traffic is dropped at the agents; 331/331 directions unchanged;
- static files carry ETags, a reload revalidates instead of pulling
  ~2 MB of vendored scripts.

KUBECONFIG is honoured for running the binary on a workstation.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-24 17:25:36 +03:00

33 lines
2.4 KiB
YAML

v0.8.0:
features:
- "Problem badges: a yellow or red \"!\" on cards, namespace and node boxes and in the tree, from object state (crash loops, pending pods, readiness, failed CronJob runs, Services without endpoints, node pressure), Warning events and Deckhouse ClusterAlerts."
- A Problems panel with every current problem, most severe first, filterable to applications or system namespaces.
- "Actions from the details panel: restart, scale, run a CronJob now, suspend and resume, delete a pod, cordon and uncordon a node. They run with the viewer's own token and RBAC; the module's service account stays read-only."
- "Load section: CPU and memory of a workload or node against requests, limits and allocatable, with a per-pod table."
- An Events tab with the events of the object, its pods and ReplicaSets.
- Cards are locked in place until the padlock is opened; the map can be saved as a PNG picture.
- The open page refreshes cards, badges and the tree every minute without moving the layout.
fixes:
- "Layout of a large namespace or node view is 3-4 times faster: levels without edges are packed without a call into ELK."
- "About 20% fewer flows through hubble-relay: the to-overlay copy of pod-to-pod traffic is filtered at the agents; traffic counts are unchanged."
- Static files are revalidated with ETags instead of downloaded again on every visit (about 2 MB).
v0.7.2:
fixes:
- A namespace without traffic in the selected window is drawn again; the empty window was sent as null and the map kept the previous namespace on screen.
- A map that fails to draw shows an error instead of the previous view.
v0.7.1:
fixes:
- Alert rules are rendered on DKP versions without the operator-prometheus-crd module.
v0.7.0:
features:
- First release as a Deckhouse module (ModuleSource, release channels, lib-helm templates).
- "Traffic collection modes: Continuous, OnDemand, Disabled; retention configurable up to 24h."
- Hubble CA is read by a module hook from d8-cni-cilium; no manual copy of the certificate.
- Authentication through user-authn (DexAuthenticator) or any external auth, as in cilium-hubble.
- English and Russian UI.
- Prometheus metrics and alerts for the Hubble stream.
fixes:
- Traffic windows now cover the requested time instead of the agent ring buffer (about 10 seconds).
- Counts are distinct connections, not Hubble events.
- Pod exits cleanly on SIGTERM.