From 2031a6f5b88173e9966933998986a73d84901c60 Mon Sep 17 00:00:00 2001 From: Kristina Pathak Date: Tue, 18 Aug 2026 15:02:02 -0700 Subject: [PATCH] fix(nvca-operator): update BYOO collector default Closes #967 Signed-off-by: Kristina Pathak --- deploy/helm/nvca-operator/nvca-operator/README.md | 2 +- .../helm/nvca-operator/nvca-operator/templates/_helpers.tpl | 2 +- deploy/helm/nvca-operator/nvca-operator/values.schema.json | 2 +- deploy/helm/nvca-operator/nvca-operator/values.yaml | 4 ++-- .../helm/nvca-operator/tests/release_image_manifest_test.sh | 6 +++--- .../tests/self_managed_nvca_image_reference_test.sh | 4 ++-- .../nvca/deployments/nvca-operator/README.md | 2 +- .../nvca/deployments/nvca-operator/templates/_helpers.tpl | 2 +- .../nvca/deployments/nvca-operator/values.schema.json | 2 +- .../nvca/deployments/nvca-operator/values.yaml | 2 +- 10 files changed, 14 insertions(+), 14 deletions(-) diff --git a/deploy/helm/nvca-operator/nvca-operator/README.md b/deploy/helm/nvca-operator/nvca-operator/README.md index 9389a3aec..9a2638911 100644 --- a/deploy/helm/nvca-operator/nvca-operator/README.md +++ b/deploy/helm/nvca-operator/nvca-operator/README.md @@ -76,7 +76,7 @@ used in Kubernetes Clusters to run NVCF Workloads. | `agent.gpuProfiling.labelKey` | Pod label key NVCA applies to profiled function pods (the label the Nsight Operator watches for). Empty uses the built-in default "nvidia-nsight-profile". | `""` | | `agent.gpuProfiling.labelValue` | Pod label value NVCA applies to profiled function pods. Empty uses the built-in default "enabled". | `""` | | `agent.byooOtelCollector.imageRepository` | Optional BYOO OpenTelemetry Collector image repository. If unset, it is derived from image.repository. | `""` | -| `agent.byooOtelCollector.imageTag` | BYOO OpenTelemetry Collector image tag. | `0.157.11` | +| `agent.byooOtelCollector.imageTag` | BYOO OpenTelemetry Collector image tag. | `0.157.0-nv-0.1.1` | | `agent.functionEnvOverrides` | Map of environment variable overrides for function workloads (e.g., {"INIT_CONTAINER": "nvcr.io/custom/init:v1.0", "UTILS_CONTAINER": "nvcr.io/custom/utils:v1.0"}) | `{}` | | `agent.taskEnvOverrides` | Map of environment variable overrides for task workloads (e.g., {"INIT_CONTAINER": "nvcr.io/custom/init:v1.0", "ESS_AGENT_CONTAINER": "nvcr.io/custom/ess:v1.0"}) | `{}` | | `agent.overrideEnvironmentVariables` | Map of environment variables to override on the NVCA agent container. These take precedence over default values. Example: {"LOG_LEVEL": "debug", "CUSTOM_FLAG": "enabled"} | `{}` | diff --git a/deploy/helm/nvca-operator/nvca-operator/templates/_helpers.tpl b/deploy/helm/nvca-operator/nvca-operator/templates/_helpers.tpl index 6b5b4f7ed..2b201f501 100644 --- a/deploy/helm/nvca-operator/nvca-operator/templates/_helpers.tpl +++ b/deploy/helm/nvca-operator/nvca-operator/templates/_helpers.tpl @@ -200,7 +200,7 @@ Get the BYOO OTel collector image when its tag is configured. {{- define "nvcaop.byooOtelCollectorImage" -}} {{- $agent := .Values.agent | default dict -}} {{- $byooOtelCollector := $agent.byooOtelCollector | default dict -}} -{{- $imageTag := "0.157.11" -}} +{{- $imageTag := "0.157.0-nv-0.1.1" -}} {{- if hasKey $byooOtelCollector "imageTag" -}} {{- $imageTag = $byooOtelCollector.imageTag -}} {{- end -}} diff --git a/deploy/helm/nvca-operator/nvca-operator/values.schema.json b/deploy/helm/nvca-operator/nvca-operator/values.schema.json index 05c451fe4..820fe026d 100644 --- a/deploy/helm/nvca-operator/nvca-operator/values.schema.json +++ b/deploy/helm/nvca-operator/nvca-operator/values.schema.json @@ -387,7 +387,7 @@ "imageTag": { "type": "string", "description": "BYOO OpenTelemetry Collector image tag.", - "default": "0.157.11" + "default": "0.157.0-nv-0.1.1" } } }, diff --git a/deploy/helm/nvca-operator/nvca-operator/values.yaml b/deploy/helm/nvca-operator/nvca-operator/values.yaml index bb6609312..be315a616 100644 --- a/deploy/helm/nvca-operator/nvca-operator/values.yaml +++ b/deploy/helm/nvca-operator/nvca-operator/values.yaml @@ -178,7 +178,7 @@ resources: ## @param agent.taskEnvOverrides Map of environment variable overrides for task workloads (e.g., {"INIT_CONTAINER": "nvcr.io/custom/init:v1.0", "ESS_AGENT_CONTAINER": "nvcr.io/custom/ess:v1.0"}) ## @param agent.overrideEnvironmentVariables Map of environment variables to override on the NVCA agent container. These take precedence over default values. Example: {"LOG_LEVEL": "debug", "CUSTOM_FLAG": "enabled"} ## @param agent.tolerations K8s tolerations for the NVCA agent pod -## @param agent.llm.requestRouterAddress Default LLM request-router worker address rendered as STARGATE_ADDRESS for LLM workers +## @param agent.llm.requestRouterAddress Operator default LLM request-router address. Workers read LLM_REQUEST_ROUTER_ADDRESS from the launch environment; not a runtime fallback ## @param agent.serviceOAuth OAuth token and JWKS endpoints used by dependent services agent: cacheMountOptionsEnabled: true @@ -203,7 +203,7 @@ agent: ## @param agent.byooOtelCollector.imageTag BYOO OpenTelemetry Collector image tag. byooOtelCollector: imageRepository: "" - imageTag: "0.157.11" + imageTag: "0.157.0-nv-0.1.1" ## Environment variable overrides for function workloads. ## Available keys: INIT_CONTAINER, UTILS_CONTAINER, OTEL_CONTAINER, BYOO_OTEL_COLLECTOR_CONTAINER, ## NICLLS_CONTAINER, ESS_AGENT_CONTAINER, INFERENCE_CONTAINER diff --git a/deploy/helm/nvca-operator/tests/release_image_manifest_test.sh b/deploy/helm/nvca-operator/tests/release_image_manifest_test.sh index 8ecf2056f..00aa8562d 100755 --- a/deploy/helm/nvca-operator/tests/release_image_manifest_test.sh +++ b/deploy/helm/nvca-operator/tests/release_image_manifest_test.sh @@ -44,9 +44,9 @@ spec: value: "nvcr.io/nvidia/nvcf-byoc/nvca" args: - --function-env-overrides-b64 - - "eyJCWU9PX09URUxfQ09MTEVDVE9SX0NPTlRBSU5FUiI6Im52Y3IuaW8vbnZpZGlhL252Y2YtYnlvYy9ieW9vLW90ZWwtY29sbGVjdG9yOjAuMTU3LjExIn0=" + - "eyJCWU9PX09URUxfQ09MTEVDVE9SX0NPTlRBSU5FUiI6Im52Y3IuaW8vbnZpZGlhL252Y2YtYnlvYy9ieW9vLW90ZWwtY29sbGVjdG9yOjAuMTU3LjAtbnYtMC4xLjEifQ==" - --task-env-overrides-b64 - - "eyJCWU9PX09URUxfQ09MTEVDVE9SX0NPTlRBSU5FUiI6Im52Y3IuaW8vbnZpZGlhL252Y2YtYnlvYy9ieW9vLW90ZWwtY29sbGVjdG9yOjAuMTU3LjExIn0=" + - "eyJCWU9PX09URUxfQ09MTEVDVE9SX0NPTlRBSU5FUiI6Im52Y3IuaW8vbnZpZGlhL252Y2YtYnlvYy9ieW9vLW90ZWwtY29sbGVjdG9yOjAuMTU3LjAtbnYtMC4xLjEifQ==" --- apiVersion: v1 kind: ConfigMap @@ -91,7 +91,7 @@ expected_images=( "nvcr.io/nvidia/nvcf-byoc/nvcf-image-credential-helper:0.5.0" "nvcr.io/nvidia/nvcf-byoc/nvcf-otel-collector:0.157.9" "nvcr.io/nvidia/nvcf-byoc/samba:1.0.5" - "nvcr.io/nvidia/nvcf-byoc/byoo-otel-collector:0.157.11" + "nvcr.io/nvidia/nvcf-byoc/byoo-otel-collector:0.157.0-nv-0.1.1" ) for image in "${expected_images[@]}"; do diff --git a/deploy/helm/nvca-operator/tests/self_managed_nvca_image_reference_test.sh b/deploy/helm/nvca-operator/tests/self_managed_nvca_image_reference_test.sh index 6026f8570..ba396fdee 100755 --- a/deploy/helm/nvca-operator/tests/self_managed_nvca_image_reference_test.sh +++ b/deploy/helm/nvca-operator/tests/self_managed_nvca_image_reference_test.sh @@ -207,12 +207,12 @@ if [[ "${byoo_function_image}" != "${byoo_task_image}" ]]; then exit 1 fi -if [[ "${byoo_function_image}" != "nvcr.io/nvidia/nvcf-byoc/byoo-otel-collector:0.157.11" ]]; then +if [[ "${byoo_function_image}" != "nvcr.io/nvidia/nvcf-byoc/byoo-otel-collector:0.157.0-nv-0.1.1" ]]; then echo "unexpected BYOO collector default: ${byoo_function_image}" >&2 exit 1 fi -if [[ "${stage_byoo_function_image}" != "stg.nvcr.io/nvidia/nvcf-byoc/byoo-otel-collector:0.157.11" || \ +if [[ "${stage_byoo_function_image}" != "stg.nvcr.io/nvidia/nvcf-byoc/byoo-otel-collector:0.157.0-nv-0.1.1" || \ "${stage_byoo_task_image}" != "${stage_byoo_function_image}" ]]; then echo "expected BYOO collector defaults to use the staging image repository" >&2 exit 1 diff --git a/src/compute-plane-services/nvca/deployments/nvca-operator/README.md b/src/compute-plane-services/nvca/deployments/nvca-operator/README.md index 20acd0894..2d79e7101 100644 --- a/src/compute-plane-services/nvca/deployments/nvca-operator/README.md +++ b/src/compute-plane-services/nvca/deployments/nvca-operator/README.md @@ -76,7 +76,7 @@ used in Kubernetes Clusters to run NVCF Workloads. | `agent.gpuProfiling.labelKey` | Pod label key NVCA applies to profiled function pods (the label the Nsight Operator watches for). Empty uses the built-in default "nvidia-nsight-profile". | `""` | | `agent.gpuProfiling.labelValue` | Pod label value NVCA applies to profiled function pods. Empty uses the built-in default "enabled". | `""` | | `agent.byooOtelCollector.imageRepository` | Optional BYOO OpenTelemetry Collector image repository. If unset, it is derived from image.repository. | `""` | -| `agent.byooOtelCollector.imageTag` | BYOO OpenTelemetry Collector image tag. | `0.157.11` | +| `agent.byooOtelCollector.imageTag` | BYOO OpenTelemetry Collector image tag. | `0.157.0-nv-0.1.1` | | `agent.functionEnvOverrides` | Map of environment variable overrides for function workloads (e.g., {"INIT_CONTAINER": "nvcr.io/custom/init:v1.0", "UTILS_CONTAINER": "nvcr.io/custom/utils:v1.0"}) | `{}` | | `agent.taskEnvOverrides` | Map of environment variable overrides for task workloads (e.g., {"INIT_CONTAINER": "nvcr.io/custom/init:v1.0", "ESS_AGENT_CONTAINER": "nvcr.io/custom/ess:v1.0"}) | `{}` | | `agent.overrideEnvironmentVariables` | Map of environment variables to override on the NVCA agent container. These take precedence over default values. Example: {"LOG_LEVEL": "debug", "CUSTOM_FLAG": "enabled"} | `{}` | diff --git a/src/compute-plane-services/nvca/deployments/nvca-operator/templates/_helpers.tpl b/src/compute-plane-services/nvca/deployments/nvca-operator/templates/_helpers.tpl index 6b5b4f7ed..2b201f501 100644 --- a/src/compute-plane-services/nvca/deployments/nvca-operator/templates/_helpers.tpl +++ b/src/compute-plane-services/nvca/deployments/nvca-operator/templates/_helpers.tpl @@ -200,7 +200,7 @@ Get the BYOO OTel collector image when its tag is configured. {{- define "nvcaop.byooOtelCollectorImage" -}} {{- $agent := .Values.agent | default dict -}} {{- $byooOtelCollector := $agent.byooOtelCollector | default dict -}} -{{- $imageTag := "0.157.11" -}} +{{- $imageTag := "0.157.0-nv-0.1.1" -}} {{- if hasKey $byooOtelCollector "imageTag" -}} {{- $imageTag = $byooOtelCollector.imageTag -}} {{- end -}} diff --git a/src/compute-plane-services/nvca/deployments/nvca-operator/values.schema.json b/src/compute-plane-services/nvca/deployments/nvca-operator/values.schema.json index 4ac4f2910..801e038cf 100644 --- a/src/compute-plane-services/nvca/deployments/nvca-operator/values.schema.json +++ b/src/compute-plane-services/nvca/deployments/nvca-operator/values.schema.json @@ -387,7 +387,7 @@ "imageTag": { "type": "string", "description": "BYOO OpenTelemetry Collector image tag.", - "default": "0.157.11" + "default": "0.157.0-nv-0.1.1" } } }, diff --git a/src/compute-plane-services/nvca/deployments/nvca-operator/values.yaml b/src/compute-plane-services/nvca/deployments/nvca-operator/values.yaml index 89da88093..bf6ba9f46 100644 --- a/src/compute-plane-services/nvca/deployments/nvca-operator/values.yaml +++ b/src/compute-plane-services/nvca/deployments/nvca-operator/values.yaml @@ -214,7 +214,7 @@ agent: ## @param agent.byooOtelCollector.imageTag BYOO OpenTelemetry Collector image tag. byooOtelCollector: imageRepository: "" - imageTag: "0.157.11" + imageTag: "0.157.0-nv-0.1.1" ## Environment variable overrides for function workloads. ## Available keys: INIT_CONTAINER, UTILS_CONTAINER, OTEL_CONTAINER, BYOO_OTEL_COLLECTOR_CONTAINER, ## NICLLS_CONTAINER, ESS_AGENT_CONTAINER, INFERENCE_CONTAINER