diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml index f69e2957f..fb76db56d 100644 --- a/.gitlab-ci.yml +++ b/.gitlab-ci.yml @@ -272,7 +272,9 @@ test: compile-services: extends: .rust-base stage: compile - needs: [test] + needs: + - job: check + optional: true # skipped in dev-release mode rules: - if: $CI_COMMIT_BRANCH == "main" && $CI_PIPELINE_SOURCE == "push" changes: diff --git a/infra/k8s/gitlab/runner-values.yaml b/infra/k8s/gitlab/runner-values.yaml index 9e23491f9..d73d5e2f5 100644 --- a/infra/k8s/gitlab/runner-values.yaml +++ b/infra/k8s/gitlab/runner-values.yaml @@ -32,11 +32,12 @@ runners: node_selector_overwrite_allowed = ".*" # L40S scale-to-zero needs ~3-5 min to provision; default 180s times out poll_timeout = 600 - # Resource limits for build pods (L40S node: ~24 vCPU, ~96GB) - cpu_request = "6000m" - cpu_limit = "20000m" - memory_request = "24Gi" - memory_limit = "80Gi" + # Default resource limits — low for Kaniko/lightweight jobs + # Rust compile/test/training jobs override via KUBERNETES_*_REQUEST CI vars + cpu_request = "500m" + cpu_limit = "4000m" + memory_request = "1Gi" + memory_limit = "8Gi" helper_cpu_request = "100m" helper_cpu_limit = "500m" helper_memory_request = "128Mi" @@ -68,8 +69,8 @@ runners: # Runner tags for job matching tags: "kapsule,rust,docker,gpu" -# Concurrency -concurrent: 4 +# Concurrency — 10 allows all 9 Kaniko builds + test to run in parallel +concurrent: 10 # RBAC for runner to spawn pods rbac: