Added Support for prometheus.remote_write queue_config (#569)

Resolves #560

charts/k8s-monitoring/README.md

@@ -233,6 +233,15 @@ The Prometheus and Loki services may be hosted on the same cluster, or remotely
 | externalServices.prometheus.protocol | string | `"remote_write"` | The type of server protocol for writing metrics. Valid options:  "remote_write" will use Prometheus Remote Write,  "otlp" will use OTLP,  "otlphttp" will use OTLP HTTP |
 | externalServices.prometheus.proxyURL | string | `""` | HTTP proxy to proxy requests to Prometheus through. |
 | externalServices.prometheus.queryEndpoint | string | `"/api/prom/api/v1/query"` | Prometheus metrics query endpoint. Preset for Grafana Cloud Metrics instances. |
+| externalServices.prometheus.queue_config.batch_send_deadline | string | 5s | Maximum time samples will wait in the buffer before sending. |
+| externalServices.prometheus.queue_config.capacity | int | 10000 | Number of samples to buffer per shard. |
+| externalServices.prometheus.queue_config.max_backoff | string | 5s | Maximum retry delay. |
+| externalServices.prometheus.queue_config.max_samples_per_send | int | 2000 | Maximum number of samples per send. |
+| externalServices.prometheus.queue_config.max_shards | int | 50 | Maximum number of concurrent shards sending samples to the endpoint. |
+| externalServices.prometheus.queue_config.min_backoff | string | 30ms | Initial retry delay. The backoff time gets doubled for each retry. |
+| externalServices.prometheus.queue_config.min_shards | int | 1 | Minimum amount of concurrent shards sending samples to the endpoint. |
+| externalServices.prometheus.queue_config.retry_on_http_429 | bool | true | Retry when an HTTP 429 status code is received. |
+| externalServices.prometheus.queue_config.sample_age_limit | string | 0s | Maximum age of samples to send. |
 | externalServices.prometheus.secret.create | bool | `true` | Should this Helm chart create the secret. If false, you must define the name and namespace values. |
 | externalServices.prometheus.secret.name | string | `""` | The name of the secret. |
 | externalServices.prometheus.secret.namespace | string | `""` | The namespace of the secret. Only used if secret.create = "false" |

charts/k8s-monitoring/templates/alloy_config/_metrics_service_remote_write.alloy.txt

@@ -24,6 +24,18 @@ prometheus.remote_write "metrics_service" {
     }
 {{- end }}
     send_native_histograms = {{ .sendNativeHistograms }}
+
+    queue_config {
+      capacity = {{ .queue_config.capacity }}
+      min_shards = {{ .queue_config.min_shards }}
+      max_shards = {{ .queue_config.max_shards }}
+      max_samples_per_send = {{ .queue_config.max_samples_per_send }}
+      batch_send_deadline = {{ .queue_config.batch_send_deadline | quote }}
+      min_backoff = {{ .queue_config.min_backoff | quote }}
+      max_backoff = {{ .queue_config.max_backoff | quote }}
+      retry_on_http_429 = {{ .queue_config.retry_on_http_429 }}
+      sample_age_limit = {{ .queue_config.sample_age_limit | quote }}
+    }
   }
 
   wal {

charts/k8s-monitoring/values.schema.json

@@ -398,6 +398,38 @@
                         "queryEndpoint": {
                             "type": "string"
                         },
+                        "queue_config": {
+                            "type": "object",
+                            "properties": {
+                                "batch_send_deadline": {
+                                    "type": "string"
+                                },
+                                "capacity": {
+                                    "type": "integer"
+                                },
+                                "max_backoff": {
+                                    "type": "string"
+                                },
+                                "max_samples_per_send": {
+                                    "type": "integer"
+                                },
+                                "max_shards": {
+                                    "type": "integer"
+                                },
+                                "min_backoff": {
+                                    "type": "string"
+                                },
+                                "min_shards": {
+                                    "type": "integer"
+                                },
+                                "retry_on_http_429": {
+                                    "type": "boolean"
+                                },
+                                "sample_age_limit": {
+                                    "type": "string"
+                                }
+                            }
+                        },
                         "secret": {
                             "type": "object",
                             "properties": {

charts/k8s-monitoring/values.yaml

@@ -77,6 +77,46 @@ externalServices:
       # @section -- External Services (Prometheus)
       passwordKey: password
 
+    # Configure the Prometheus Remote Write Queue
+    # [docs](https://grafana.com/docs/alloy/latest/reference/components/prometheus.remote_write/#queue_config-block)
+    queue_config:
+      # -- Number of samples to buffer per shard.
+      # @default -- 10000
+      # @section -- External Services (Prometheus)
+      capacity: 10000
+      # -- Minimum amount of concurrent shards sending samples to the endpoint.
+      # @default -- 1
+      # @section -- External Services (Prometheus)
+      min_shards: 1
+      # -- Maximum number of concurrent shards sending samples to the endpoint.
+      # @default -- 50
+      # @section -- External Services (Prometheus)
+      max_shards: 50
+      # -- Maximum number of samples per send.
+      # @default -- 2000
+      # @section -- External Services (Prometheus)
+      max_samples_per_send: 2000
+      # -- Maximum time samples will wait in the buffer before sending.
+      # @default -- 5s
+      # @section -- External Services (Prometheus)
+      batch_send_deadline: 5s
+      # -- Initial retry delay. The backoff time gets doubled for each retry.
+      # @default -- 30ms
+      # @section -- External Services (Prometheus)
+      min_backoff: 30ms
+      # -- Maximum retry delay.
+      # @default -- 5s
+      # @section -- External Services (Prometheus)
+      max_backoff: 5s
+      # -- Retry when an HTTP 429 status code is received.
+      # @default -- true
+      # @section -- External Services (Prometheus)
+      retry_on_http_429: true
+      # -- Maximum age of samples to send.
+      # @default -- 0s
+      # @section -- External Services (Prometheus)
+      sample_age_limit: 0s
+
     # Credential management
     secret:
       # -- Should this Helm chart create the secret. If false, you must define the name and namespace values.