feat(observability): turn on the mcp-kubernetes chart's own monitor and boards - #618
Merged
Merged
Conversation
QuentinBisson
marked this pull request as ready for review
September 22, 2026 09:34
Published Helm chart
|
| Chart | agent-platform |
| Version | 4.52.1-dev.feat-mcp-k--monitoring.2026-09-22.10-02-21.hdd194fa |
| OCI reference | oci://gsoci.azurecr.io/charts/giantswarm/agent-platform:4.52.1-dev.feat-mcp-k--monitoring.2026-09-22.10-02-21.hdd194fa |
| Digest | sha256:3af07e9435ccf9ab4b517e94a12ff80ed1eebc73f671fc1798ec1441603dcefd |
| Registry | public — gsoci.azurecr.io |
| Git catalog | giantswarm-test-catalog (index) |
Pull this chart
helm pull oci://gsoci.azurecr.io/charts/giantswarm/agent-platform --version 4.52.1-dev.feat-mcp-k--monitoring.2026-09-22.10-02-21.hdd194faPosted by architect-orb · build 11069 · commit dd194fa · updated in place on every push
Published Helm chart
|
| Chart | agent-platform-connectivity |
| Version | 4.52.1-dev.feat-mcp-k--monitoring.2026-09-22.10-02-21.hdd194fa |
| OCI reference | oci://gsoci.azurecr.io/charts/giantswarm/agent-platform-connectivity:4.52.1-dev.feat-mcp-k--monitoring.2026-09-22.10-02-21.hdd194fa |
| Digest | sha256:cad70ea3d0198d7608021178fe4809852424cede9e47ca844de09bc87539d047 |
| Registry | public — gsoci.azurecr.io |
| Git catalog | giantswarm-test-catalog (index) |
Pull this chart
helm pull oci://gsoci.azurecr.io/charts/giantswarm/agent-platform-connectivity --version 4.52.1-dev.feat-mcp-k--monitoring.2026-09-22.10-02-21.hdd194faPosted by architect-orb · build 11068 · commit dd194fa · updated in place on every push
…nd boards The chart ships a ServiceMonitor over its dedicated metrics port and three Grafana boards, and this chart turned none of them on. mcp-kubernetes.mcpKubernetes.instrumentation.serviceMonitor.enabled and mcp-kubernetes.grafanaDashboards.enabled follow the resolved global.observability.metrics.serviceMonitor.enabled; the monitor carries the tenant label and the boards land in Shared Org / Agent Platform. prometheusRules stay off: the three alerts link to runbook pages that do not exist yet.
QuentinBisson
force-pushed
the
feat/mcp-kubernetes-monitoring
branch
from
September 22, 2026 10:02
691ebe6 to
dd194fa
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Problem
giantswarm/mcp-kubernetesshipstemplates/servicemonitor.yamlover its dedicated metrics port and three Grafana boards (administrator,security,cluster-operator), all behind switches that default to off. This chart set none of them, so an installation that runs the Kubernetes MCP server collects nothing from it: no scrape, no board, and the tool-call, operation and auth-failure series its instrumentation records are served and thrown away.Change
Two leaves in the
mcp-kubernetesblock, bothautoand both following the resolvedglobal.observability.metrics.serviceMonitor.enabledlike every other monitor of this chart:mcpKubernetes.instrumentation.serviceMonitor.enabled, withobservability.giantswarm.io/tenant: giantswarmon the monitor. Without that label Mimir routes the scrape to no tenant.grafanaDashboards.enabled, withfolder: Agent Platformandgiantswarm.organization: Shared Org, so the three boards land beside the platform's own instead of the chart's default folder named after the component.mcpKubernetes.metrics.enabledandmcpKubernetes.instrumentation.enabledare already true in the chart, so the endpoint needs nothing here.prometheusRulesstays off. The chart's three alerts link tomcp-kubernetes-high-error-rate,mcp-kubernetes-operation-failuresandmcp-kubernetes-auth-failures, and none of those runbook pages exists ingiantswarm/giantswarm. Turning the rules on would page on-call with three dead links. The runbooks come first, in their own pull request.No range moves
The floor
1.1.1already renders both: pulled 1.1.1, 1.1.8 (the BOM pin) and 1.2.23 (what the range resolves) and templated each with these values. All three render one ServiceMonitor and three dashboard ConfigMaps, each carryingobservability.giantswarm.io/organization.Tests
verify-target: the new block is held on both sides of the golden comparison, and dropped onceGOLDEN_REFcarries it.verify-components-charts: the chart accepts the forwarded values at the range resolution and at the BOM pin.verify-modes,verify-componentsand the pre-commit hooks (schema, helm-docs) pass.Also restores the
## [Unreleased]heading inCHANGELOG.md. A sed replacement in #609 collapsed it into a literal\1, and #615 restored only the### Addedbelow it.Part of giantswarm/giantswarm#36711.