Follow up on NetworkPolicy review nits from PR #2494
I maintainer di solito rispondono entro 1 giorno
@kaovilai ci sta già lavorando.
Dal 16/10/2026.
Valutazione
Questa issue non è ancora stata valutata.
Descrizione
Summary
Track the actionable unresolved inline review nits from PR #2494, which scopes operand NetworkPolicies and fixes DNS egress.
Requested by @kaovilai in https://github.com/openshift/oadp-operator/pull/2494#issuecomment-6223577628.
Source PR: https://github.com/openshift/oadp-operator/pull/2494.
Required changes
1. Report unexpected mover pod watch closure
Affected area: tests/e2e/networkpolicy_movers_test.go, in the goroutine that consumes events.ResultChan().
The review identifies a silent exit when the watch channel closes. If the API server closes the watch during a backup or restore, later mover pods are not observed. Verification can then report a missing mover pod instead of the watch failure.
Set watchErr when the channel closes while the context remains active, unless an error is already recorded. Protect shared state with the existing mutex. Do not report normal context cancellation and cleanup as failures. A retrying watch or informer is an optional resilience improvement, not a prerequisite for this diagnostic fix.
Review references:
- https://github.com/openshift/oadp-operator/pull/2494#discussion_r4287964896
- https://github.com/openshift/oadp-operator/pull/2494#discussion_r4291337367
Acceptance criteria:
- Unexpected channel closure produces a specific watch-closure diagnostic.
- Normal cancellation and cleanup do not produce a watch error.
- An existing watch error is preserved.
- Focused tests cover unexpected closure, normal cancellation, and existing-error preservation.
2. Document the legacy default-deny stale-owner migration edge case
Affected areas: docs/network_policies.md and the behavior of deleteLegacyDefaultDenyNetworkPolicy in internal/controller/networkpolicy.go.
The review identifies an edge case during rapid DPA deletion and recreation in the same namespace. The replacement DPA has a different UID. If garbage collection has not yet removed the old DPA's legacy default-deny policy, metav1.IsControlledBy does not identify that policy as controlled by the replacement DPA. Reconciliation preserves the policy, which can continue namespace-wide isolation until garbage collection removes it.
Document this migration limitation, its effect on partner controllers, and safe troubleshooting guidance for delayed garbage collection. Preserve the current ownership safeguards. Any proposal to delete policies controlled by a different DPA UID requires a separate safety evaluation.
Review reference:
Acceptance criteria:
- The migration documentation describes the rapid delete/recreate scenario and UID mismatch.
- The documentation explains that the legacy policy can continue isolating unrelated pods until garbage collection completes.
- Troubleshooting guidance requires checking policy ownership and DPA state before manual intervention.
- The follow-up does not weaken ownership checks or UID/resource-version deletion preconditions.
Scope
The two mover-watch review threads describe one work item. The inline comment that confirms empty-ingress behavior does not require a change. This issue tracks the actionable inline nits; it does not reclassify or resolve the separate automated pre-merge check findings.
- Lingua principale
- Go
- Stelle
- 93
- Fork
- 93
- Merge medio
- 1g 20h
- PR unite (30g)
- 27
Preparare l'ambiente
- Include un Dockerfile o un file Docker Compose
- Ha un modello di pull request
- Leggi la guida per i contributori
Come iniziare
- Leggi tutta la issue e poi la guida ai contributi del progetto.
- Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
- Fai un fork del repository e lavora su un branch.
- Apri una pull request che faccia riferimento al numero della issue.
Altre issue di openshift/oadp-operator
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 78/100
openshift/oadp-operator#2372 ·
I maintainer di solito rispondono entro 1 giorno
-
Makefile container tool check only verifies binary exists, not runtime readinessForse già presa @Shreyashxredhat l’ha presa 116 giorni fa. Apertalifecycle/rotten
Difficoltà 2/5 1-3 ore Idoneità per principianti 78/100
openshift/oadp-operator#2249 · 2 commenti ·
I maintainer di solito rispondono entro 1 giorno
-
oadp-1.4 periodic e2e (4.18 AWS): "Multiple BSL with custom CA cert" times out on "Default BSL should be available"Forse già presa Una pull request collegata a questa issue è aperta o già unita. Aperta
Difficoltà 4/5 3-5 giorni Idoneità per principianti 48/100
openshift/oadp-operator#2514 ·
I maintainer di solito rispondono entro 1 giorno
-
tide/merge-blocker
Difficoltà 1/5 Meno di un'ora Idoneità per principianti 10/100
openshift/oadp-operator#2501 ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 4/5 1-2 giorni Idoneità per principianti 15/100
openshift/oadp-operator#2490 ·
I maintainer di solito rispondono entro 1 giorno
Tutte le issue di openshift/oadp-operator
Issue simili
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 75/100
github/github-mcp-server#3581 ·
I maintainer di solito rispondono entro 2 giorni
-
agent-review-finding chore
Difficoltà 1/5 Meno di un'ora Idoneità per principianti 85/100
jordansmall/spindrift#6033 ·
I maintainer di solito rispondono entro 1 giorno
-
Replace UUID implementationApertaenhancement
Difficoltà 2/5 1-3 ore Idoneità per principianti 75/100
hyperledger/fabric-x-common#230 ·
I maintainer di solito rispondono entro 1 giorno
-
About.md no longer usedApertabug
Difficoltà 2/5 Mezza giornata Idoneità per principianti 65/100
open-wanderer/wanderer#1443 · 3 commenti ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 75/100
I maintainer di solito rispondono entro 1 giorno