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>
204 lines
8.3 KiB
YAML
204 lines
8.3 KiB
YAML
type: object
|
||
properties:
|
||
traffic:
|
||
type: object
|
||
default: {}
|
||
description: Live traffic between workloads, read from Cilium Hubble Relay.
|
||
properties:
|
||
mode:
|
||
type: string
|
||
default: Continuous
|
||
enum: ["Continuous", "OnDemand", "Disabled"]
|
||
description: |
|
||
When flows are read from Hubble.
|
||
|
||
- `Continuous` — a Hubble follow stream is always open, so every traffic window is full as soon as the map is opened. Costs hubble-relay about 100–150m CPU on a busy cluster.
|
||
- `OnDemand` — the stream opens with the first traffic request and closes after `onDemandIdleMinutes` without one. Near-zero cost while nobody looks at the map, but windows only fill while it is in use.
|
||
- `Disabled` — no connection to Hubble; the map shows configuration links only.
|
||
|
||
Traffic also stays off when the `cilium-hubble` module is disabled.
|
||
retentionMinutes:
|
||
type: integer
|
||
default: 60
|
||
minimum: 5
|
||
maximum: 1440
|
||
description: |
|
||
How far back traffic windows can reach, in minutes. Memory grows with retention and with the number of distinct connections in the cluster.
|
||
onDemandIdleMinutes:
|
||
type: integer
|
||
default: 10
|
||
minimum: 1
|
||
maximum: 120
|
||
description: In `OnDemand` mode, how long the stream stays open after the last traffic request.
|
||
collectInfraTraffic:
|
||
type: boolean
|
||
default: false
|
||
description: |
|
||
Also collect kubelet probes and other traffic originating from node hosts, node-to-node traffic, Cilium health checks and DNS.
|
||
|
||
Off by default: on a typical cluster this is about three quarters of all flows, none of it maps to a workload, and filtering it at the Hubble agents keeps hubble-relay load several times lower.
|
||
actions:
|
||
type: object
|
||
default: {}
|
||
description: Changes the map can make to the cluster.
|
||
properties:
|
||
enabled:
|
||
type: boolean
|
||
default: true
|
||
description: |
|
||
Buttons to restart a workload, scale a Deployment or StatefulSet, delete a pod, run a CronJob now or suspend it, cordon a node.
|
||
|
||
Every action runs with the viewer's own Kubernetes identity: the DexAuthenticator passes the user's token to the module, and the API server checks it against the user's RBAC. The module's service account stays read-only, a button is shown only when the user's role allows it, and the API server's audit log records the user.
|
||
|
||
Needs the `user-authn` module, or an `externalAuthentication` that puts a token the API server accepts into the `Authorization` header. When disabled, the user's token is not passed to the module at all.
|
||
logs:
|
||
type: object
|
||
default: {}
|
||
properties:
|
||
tailLines:
|
||
type: integer
|
||
default: 500
|
||
minimum: 10
|
||
maximum: 10000
|
||
description: How many last lines of a container log the Logs tab shows.
|
||
systemNamespacePrefixes:
|
||
type: array
|
||
default: ["d8-", "kube-"]
|
||
items:
|
||
type: string
|
||
description: |
|
||
Namespaces grouped as "System" in the inventory tree. An entry ending in `-` matches by prefix, any other entry matches the whole name.
|
||
ingressClass:
|
||
type: string
|
||
pattern: '^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$'
|
||
description: |
|
||
The class of the Ingress controller used for the web interface.
|
||
|
||
Optional. By default, the `modules.ingressClass` global value is used.
|
||
nodeSelector:
|
||
type: object
|
||
additionalProperties:
|
||
type: string
|
||
description: |
|
||
The same as the `spec.nodeSelector` pod parameter in Kubernetes.
|
||
|
||
If the parameter is omitted or `false`, it will be determined [automatically](https://deckhouse.io/products/kubernetes-platform/documentation/v1/#advanced-scheduling).
|
||
tolerations:
|
||
type: array
|
||
description: |
|
||
The same as `spec.tolerations` for the Kubernetes Pod.
|
||
|
||
If the parameter is omitted or `false`, it will be determined [automatically](https://deckhouse.io/products/kubernetes-platform/documentation/v1/#advanced-scheduling).
|
||
items:
|
||
type: object
|
||
properties:
|
||
effect:
|
||
type: string
|
||
key:
|
||
type: string
|
||
operator:
|
||
type: string
|
||
tolerationSeconds:
|
||
type: integer
|
||
format: int64
|
||
value:
|
||
type: string
|
||
auth:
|
||
type: object
|
||
default: {}
|
||
description: |
|
||
Access to the web interface.
|
||
|
||
Anyone admitted here sees every namespace, the manifests of workloads, Services and Ingresses (never Secrets) and container logs, through the module's own read-only service account. Restrict access with `allowedUserGroups`.
|
||
properties:
|
||
externalAuthentication:
|
||
type: object
|
||
description: |
|
||
Parameters to enable external authentication based on the Ingress NGINX [external-auth](https://kubernetes.github.io/ingress-nginx/examples/auth/external-auth/) mechanism.
|
||
|
||
> External authentication is enabled automatically if the [user-authn](https://deckhouse.io/modules/user-authn/) module is enabled.
|
||
properties:
|
||
authURL:
|
||
type: string
|
||
x-examples: ["https://example.com/dex/auth"]
|
||
description: The URL of the authentication service. If the user is authenticated, the service should return an HTTP 200 response code.
|
||
authSignInURL:
|
||
type: string
|
||
x-examples: ["https://example.com/dex/sign_in"]
|
||
description: The URL to redirect the user for authentication (if the authentication service returned a non-200 HTTP response code).
|
||
allowedUserEmails:
|
||
type: array
|
||
items:
|
||
type: string
|
||
description: |
|
||
Emails of users allowed to open the web interface.
|
||
|
||
Used if the `user-authn` module is enabled or `externalAuthentication` is set.
|
||
allowedUserGroups:
|
||
type: array
|
||
items:
|
||
type: string
|
||
x-examples:
|
||
- ["admins"]
|
||
description: |
|
||
User groups allowed to open the web interface.
|
||
|
||
Used if the `user-authn` module is enabled or `externalAuthentication` is set.
|
||
|
||
**Caution!** When used together with the [user-authn](https://deckhouse.io/modules/user-authn/) module, add these groups to the matching field of the DexProvider.
|
||
whitelistSourceRanges:
|
||
type: array
|
||
items:
|
||
type: string
|
||
x-examples:
|
||
- ["1.1.1.1/32"]
|
||
description: CIDRs allowed to open the web interface.
|
||
https:
|
||
type: object
|
||
x-examples:
|
||
- mode: Disabled
|
||
- mode: OnlyInURI
|
||
- mode: CustomCertificate
|
||
customCertificate:
|
||
secretName: "foobar"
|
||
- mode: CertManager
|
||
certManager:
|
||
clusterIssuerName: letsencrypt
|
||
description: |
|
||
What certificate type to use.
|
||
|
||
This parameter completely overrides the `global.modules.https` settings.
|
||
properties:
|
||
mode:
|
||
type: string
|
||
default: "CertManager"
|
||
description: |
|
||
The HTTPS usage mode:
|
||
- `CertManager` — the web interface is served over HTTPS with a certificate from the ClusterIssuer set in `certManager.clusterIssuerName`;
|
||
- `CustomCertificate` — the web interface is served over HTTPS with a certificate from the `d8-system` namespace;
|
||
- `Disabled` — the web interface is served over HTTP only;
|
||
- `OnlyInURI` — the web interface is served over HTTP behind an external HTTPS load balancer that terminates TLS; links are generated with the HTTPS scheme.
|
||
enum:
|
||
- "Disabled"
|
||
- "CertManager"
|
||
- "CustomCertificate"
|
||
- "OnlyInURI"
|
||
certManager:
|
||
type: object
|
||
properties:
|
||
clusterIssuerName:
|
||
type: string
|
||
default: "letsencrypt"
|
||
x-examples: ["letsencrypt", "letsencrypt-staging", "selfsigned"]
|
||
description: The ClusterIssuer used to get the certificate.
|
||
customCertificate:
|
||
type: object
|
||
default: {}
|
||
properties:
|
||
secretName:
|
||
type: string
|
||
description: |
|
||
The name of the Secret in the `d8-system` namespace to use for the web interface.
|
||
|
||
The Secret must have the [kubernetes.io/tls](https://kubernetes.github.io/ingress-nginx/user-guide/tls/#tls-secrets) format.
|