feat(infra): remove leftover Envoy Gateway, complete Cilium cutover - #31
Open
jonathandieu wants to merge 1 commit into
Open
feat(infra): remove leftover Envoy Gateway, complete Cilium cutover#31jonathandieu wants to merge 1 commit into
jonathandieu wants to merge 1 commit into
Conversation
PR #29 added the Cilium Gateway but left Envoy Gateway deployed alongside it. All HTTPRoutes already parent onto the Cilium gateway (cilium-gateway-system/public), so Envoy serves no traffic; its orphaned DigitalOcean LoadBalancer is what drives the CCM's perpetual "422 already a load balancer with that name" reconcile loop. Remove Envoy Gateway entirely: - delete charts/infrastructure/envoy-gateway - delete argocd/apps/infrastructure/envoy-gateway-appset.yaml - delete clusters/{platform,stage}-do-atl1/overrides/envoy-gateway.yaml - repoint the two comment-only references (cert-manager, plausible) to Cilium The root app-of-apps prunes the ApplicationSet on merge and the generated Applications carry resources-finalizer, so Envoy resources (incl. the LB) cascade-delete. gateway-api-crds and cert-manager's enableGatewayAPI stay; Cilium relies on both.
Helm template diffOnly in /tmp/base: envoy-gateway.yaml
diff -u --recursive --label base --label head base head
--- base
+++ head
@@ -218,7 +218,7 @@
app.kubernetes.io/managed-by: Helm
type: Opaque
data:
- SECRET_KEY_BASE: TkUzNk80RXBlVngzdkEwU0RRUVZ5a2EyUnF2dXhySGc5ekpaYjZnZnhYMW1QT1RLcVM4aXo4SGNsNWJWeFV3NkFNU1VrNlhjUmxXWndEMFh4Ukl6SVFLSml2
+ SECRET_KEY_BASE: dldxQ2psWHdzZXZZVDh4WkZkTzFRSFFGOEtCYUozVEFabTJrWGJkSFcxazNsZkh2bVVFNnp4azV5MTE2VWk5RkZZZU80Ym5kWWVzMkxrVGVNNXFPZWdMVG1N
TOTP_VAULT_KEY: ZHN4dmJuM2p4RGQxNmF6MlFwc1g1QjhPK2xseGpRMlNKRTJpNUJ6eDM4ST0=
DATABASE_URL: cG9zdGdyZXM6Ly9wb3N0Z3Jlczpwb3N0Z3Jlc0BwbGF1c2libGUtcG9zdGdyZXNxbDo1NDMyL3BsYXVzaWJsZV9kYg==
CLICKHOUSE_DATABASE_URL: aHR0cDovL2NsaWNraG91c2U6cGFzc3dvcmRAcGxhdXNpYmxlLWNsaWNraG91c2U6ODEyMy9wbGF1c2libGVfZXZlbnRzX2Ri |
jonathandieu
marked this pull request as ready for review
August 7, 2026 02:08
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
1 participant
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.
Summary
PR #29 introduced the Cilium Gateway but left Envoy Gateway deployed alongside it. All three HTTPRoutes already parent onto the Cilium gateway (
cilium-gateway-system/public), so Envoy serves no traffic. Its orphaned DigitalOcean LoadBalancer is what drives the cloud-controller's perpetual422 There is already a load balancer with that namereconcile loop (~1300 failures over ~4.5 days), which also keeps external-dns from settling.This removes Envoy Gateway entirely so Cilium is the sole gateway.
Changes
charts/infrastructure/envoy-gatewaychartargocd/apps/infrastructure/envoy-gateway-appset.yamlclusters/{platform,stage}-do-atl1/overrides/envoy-gateway.yamlcert-manager,plausiblevalues) to CiliumNo functional Envoy references remain.
gateway-api-crdsand cert-manager'senableGatewayAPIstay; Cilium relies on both.Verification
ciliumGatewayClass Accepted, Gatewaycilium-gateway-system/publicPROGRAMMED (LB129.212.241.198), adopting the terraform LB (load-balancer-id f0f43b4a..., typeREGIONAL). Hitting it directly,recipesandchangesreturn 200,analytics503 (Plausible app-level).parentRefthe Cilium gateway.0d90b7c6..., typeREGIONAL_NETWORK) and was the source of the 422 churn.Cleanup on merge
The
rootapp-of-apps hasprune: trueand the Envoy-generated Applications carryresources-finalizer.argocd.argoproj.io, so on merge theenvoy-gatewayApplicationSet is pruned and its resources (including the DO LoadBalancer) cascade-delete.Post-merge, verify:
kubectl -n argocd get applicationsetshows noenvoy-gatewayenvoy-gateway-systemnamespace and its DO LB are goneDraft until the post-merge cleanup path is confirmed acceptable.