diff --git a/.forgejo/workflows/ci.yml b/.forgejo/workflows/ci.yml index ed6ca70f4..b0a5840b8 100644 --- a/.forgejo/workflows/ci.yml +++ b/.forgejo/workflows/ci.yml @@ -71,15 +71,6 @@ jobs: # nix-shell-for-one-tool shape the vale jobs below use. run: nix shell --inputs-from . nixpkgs#jq --command sh scripts/check-dashboard-descriptions.sh - dashboard-service-names: - name: dashboard-service-name lint - if: vars.PUBLIC_FORGE != 'true' - runs-on: ${{ vars.PUBLIC_FORGE == 'true' && 'nixos' || 'hive-ci' }} - steps: - - uses: actions/checkout@v3 - - name: lint - run: sh scripts/check-dashboard-service-names.sh - shellcheck: name: shellcheck if: vars.PUBLIC_FORGE != 'true' diff --git a/docs/swarm/services.md b/docs/swarm/services.md index 406461a59..7cb0f4ee1 100644 --- a/docs/swarm/services.md +++ b/docs/swarm/services.md @@ -207,7 +207,7 @@ kept. - `review-classifier`: review requests by outcome, the verdict mix, results per check, retried asks, and run duration p50 and p95. - `forge`: what the forge holds, and the forge process's CPU, memory and descriptors. - `gateway`: the gateway nginx on each hive's host: requests per second, connections by state, and accepted against handled connections. -- `collector`: what the hive and swarm OTEL collectors accept, refuse, export and queue, with queue depth against capacity. +- `collector`: what the OTEL collectors accept, refuse, export and queue, with queue depth against capacity. - `logstore`, `metricstore`: health of the log store and the metrics store. A new panel can read "No data" until its source exists: the push and commit @@ -215,17 +215,6 @@ counters on `controller` appear after the first push webhook, the refused and failure series on `collector` appear once something fails, and the `review-classifier` panels fill as review requests arrive. -**The `collector` dashboard.** It shows the self-metrics of the hive's -collector and of the swarm's collector, which both carry the collector -binary's default `service.name`, `otelcol-contrib`. The timeseries panels -split by `server.port`: 8888 is the hive collector, 8889 the swarm -collector. The per-container collectors (the secret store, NATS, Grafana, -Authelia, the Victoria stores and the agents' forwarders) don't report -their own metrics yet, so they have no lines here; a separate issue on -the forge covers that. `scripts/check-dashboard-service-names.sh` runs in CI -and in `scripts/pre-push`. It fails when a dashboard query selects a -`service.name` that nothing emits. - **Where the data comes from.** The swarm's OTEL collector, below. The `controller` panels come from the controller itself. It exports OTLP diff --git a/nix/host-modules/swarm-grafana/dashboards/collector.json b/nix/host-modules/swarm-grafana/dashboards/collector.json index 8f1facb57..b2b720df4 100644 --- a/nix/host-modules/swarm-grafana/dashboards/collector.json +++ b/nix/host-modules/swarm-grafana/dashboards/collector.json @@ -1,7 +1,7 @@ { "title": "hyperhive \u00b7 collector (self-telemetry)", "uid": "hyperhive-swarm-collector", - "description": "What the OTEL collectors accept, refuse, export and queue. Port 8888 is the hive, 8889 swarm-otel.", + "description": "What the OTEL collectors accept, refuse, export and queue.", "editable": true, "refresh": "1m", "schemaVersion": 39, @@ -35,7 +35,7 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "sum(increase({__name__=~\"otelcol_receiver_(refused|failed)_(metric_points|log_records|spans)(_total)?\", \"service.name\"=\"otelcol-contrib\"}[$__range])) or vector(0)", + "expr": "sum(increase({__name__=~\"otelcol_receiver_(refused|failed)_(metric_points|log_records|spans)(_total)?\", \"service.name\"=\"collector\"}[$__range])) or vector(0)", "instant": true, "legendFormat": "refused" } @@ -99,7 +99,7 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "sum(increase({__name__=~\"otelcol_exporter_(send|enqueue)_failed_(metric_points|log_records|spans)(_total)?\", \"service.name\"=\"otelcol-contrib\"}[$__range])) or vector(0)", + "expr": "sum(increase({__name__=~\"otelcol_exporter_(send|enqueue)_failed_(metric_points|log_records|spans)(_total)?\", \"service.name\"=\"collector\"}[$__range])) or vector(0)", "instant": true, "legendFormat": "failed" } @@ -163,7 +163,7 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "max(otelcol_exporter_queue_size{\"service.name\"=\"otelcol-contrib\"} / otelcol_exporter_queue_capacity{\"service.name\"=\"otelcol-contrib\"})", + "expr": "max(otelcol_exporter_queue_size{\"service.name\"=\"collector\"} / otelcol_exporter_queue_capacity{\"service.name\"=\"collector\"})", "instant": true, "legendFormat": "fill" } @@ -230,7 +230,7 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "max({__name__=~\"otelcol_process_memory_rss(_bytes)?\", \"service.name\"=\"otelcol-contrib\"})", + "expr": "max({__name__=~\"otelcol_process_memory_rss(_bytes)?\", \"service.name\"=\"collector\"})", "instant": true, "legendFormat": "rss" } @@ -289,8 +289,8 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "sum by (\"server.port\") (rate({__name__=~\"otelcol_receiver_accepted_metric_points(_total)?\", \"service.name\"=\"otelcol-contrib\"}[$__rate_interval]))", - "legendFormat": "{{server.port}} metric points" + "expr": "sum(rate({__name__=~\"otelcol_receiver_accepted_metric_points(_total)?\", \"service.name\"=\"collector\"}[$__rate_interval]))", + "legendFormat": "metric points" }, { "refId": "B", @@ -298,8 +298,8 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "sum by (\"server.port\") (rate({__name__=~\"otelcol_receiver_accepted_log_records(_total)?\", \"service.name\"=\"otelcol-contrib\"}[$__rate_interval]))", - "legendFormat": "{{server.port}} log records" + "expr": "sum(rate({__name__=~\"otelcol_receiver_accepted_log_records(_total)?\", \"service.name\"=\"collector\"}[$__rate_interval]))", + "legendFormat": "log records" }, { "refId": "C", @@ -307,8 +307,8 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "sum by (\"server.port\") (rate({__name__=~\"otelcol_receiver_accepted_spans(_total)?\", \"service.name\"=\"otelcol-contrib\"}[$__rate_interval]))", - "legendFormat": "{{server.port}} spans" + "expr": "sum(rate({__name__=~\"otelcol_receiver_accepted_spans(_total)?\", \"service.name\"=\"collector\"}[$__rate_interval]))", + "legendFormat": "spans" } ], "fieldConfig": { @@ -364,8 +364,8 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "sum by (receiver, \"server.port\") (rate({__name__=~\"otelcol_receiver_(refused|failed)_(metric_points|log_records|spans)(_total)?\", \"service.name\"=\"otelcol-contrib\"}[$__rate_interval]))", - "legendFormat": "{{server.port}} {{receiver}}" + "expr": "sum by (receiver) (rate({__name__=~\"otelcol_receiver_(refused|failed)_(metric_points|log_records|spans)(_total)?\", \"service.name\"=\"collector\"}[$__rate_interval]))", + "legendFormat": "{{receiver}}" } ], "fieldConfig": { @@ -421,8 +421,8 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "sum by (exporter, \"server.port\") (rate({__name__=~\"otelcol_exporter_sent_(metric_points|log_records|spans)(_total)?\", \"service.name\"=\"otelcol-contrib\"}[$__rate_interval]))", - "legendFormat": "{{server.port}} {{exporter}}" + "expr": "sum by (exporter) (rate({__name__=~\"otelcol_exporter_sent_(metric_points|log_records|spans)(_total)?\", \"service.name\"=\"collector\"}[$__rate_interval]))", + "legendFormat": "{{exporter}}" } ], "fieldConfig": { @@ -478,8 +478,8 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "sum by (exporter, \"server.port\") (rate({__name__=~\"otelcol_exporter_(send|enqueue)_failed_(metric_points|log_records|spans)(_total)?\", \"service.name\"=\"otelcol-contrib\"}[$__rate_interval]))", - "legendFormat": "{{server.port}} {{exporter}}" + "expr": "sum by (exporter) (rate({__name__=~\"otelcol_exporter_(send|enqueue)_failed_(metric_points|log_records|spans)(_total)?\", \"service.name\"=\"collector\"}[$__rate_interval]))", + "legendFormat": "{{exporter}}" } ], "fieldConfig": { @@ -535,8 +535,8 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "max by (exporter, \"server.port\") (otelcol_exporter_queue_size{\"service.name\"=\"otelcol-contrib\"})", - "legendFormat": "{{server.port}} {{exporter}} size" + "expr": "max by (exporter) (otelcol_exporter_queue_size{\"service.name\"=\"collector\"})", + "legendFormat": "{{exporter}} size" }, { "refId": "B", @@ -544,8 +544,8 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "max by (exporter, \"server.port\") (otelcol_exporter_queue_capacity{\"service.name\"=\"otelcol-contrib\"})", - "legendFormat": "{{server.port}} {{exporter}} capacity" + "expr": "max by (exporter) (otelcol_exporter_queue_capacity{\"service.name\"=\"collector\"})", + "legendFormat": "{{exporter}} capacity" } ], "fieldConfig": { @@ -600,8 +600,8 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "max by (\"server.port\") ({__name__=~\"otelcol_process_memory_rss(_bytes)?\", \"service.name\"=\"otelcol-contrib\"})", - "legendFormat": "{{server.port}} rss" + "expr": "max({__name__=~\"otelcol_process_memory_rss(_bytes)?\", \"service.name\"=\"collector\"})", + "legendFormat": "rss" }, { "refId": "B", @@ -609,8 +609,8 @@ "type": "prometheus", "uid": "@datasourceUid@" }, - "expr": "max by (\"server.port\") (otelcol_process_runtime_heap_alloc_bytes{\"service.name\"=\"otelcol-contrib\"})", - "legendFormat": "{{server.port}} heap" + "expr": "max(otelcol_process_runtime_heap_alloc_bytes{\"service.name\"=\"collector\"})", + "legendFormat": "heap" } ], "fieldConfig": { diff --git a/scripts/check-dashboard-service-names.sh b/scripts/check-dashboard-service-names.sh deleted file mode 100644 index bd365fc65..000000000 --- a/scripts/check-dashboard-service-names.sh +++ /dev/null @@ -1,67 +0,0 @@ -#!/bin/sh -# CI lint: every literal `"service.name"=""` selector in a shipped grafana -# dashboard must name a service.name that something emits. A dashboard -# querying a name nobody sets renders empty panels with no error. -# -# `known` lists the emitted names, one `name|file|anchor` per line. The -# anchor is a fixed string that must still appear in that file, so an entry -# cannot outlive the code that sets the name. Regex selectors -# (`"service.name"=~"$service"`) are template variables and are not checked. -# -set -eu - -dashboards_dir="nix/host-modules/swarm-grafana/dashboards" - -# bao: `collectorServiceName` upserted on the store's collector. -# authelia, forgejo: `serviceName` of the journal forwarder in each module. -# swarm-controller: the resource of the controller's own metrics. -# otelcol-contrib: the unset default of the collector binary, which the hive -# collector and swarm-otel both run (`pkgs.opentelemetry-collector-contrib`) -# and which stamps their self-metrics. -known='bao|nix/host-modules/swarm-bao.nix|collectorServiceName = "bao" -authelia|nix/host-modules/swarm-authelia.nix|serviceName = "authelia" -forgejo|nix/host-modules/hive-forge/default.nix|serviceName = "forgejo" -swarm-controller|swarm-controller/src/vcs_metrics.rs|with_service_name("swarm-controller") -otelcol-contrib|nix/host-modules/otel.nix|package = pkgs.opentelemetry-collector-contrib' - -violations=0 - -while IFS='|' read -r name file anchor; do - if ! grep -qF -- "$anchor" "$file" 2>/dev/null; then - violations=$((violations + 1)) - printf '::error file=%s::service.name "%s" is listed as emitted but %s no longer contains: %s\n' \ - "$file" "$name" "$file" "$anchor" - fi -done <&2 - exit 1 -fi -if [ "$violations" -gt 0 ]; then - printf 'check-dashboard-service-names: %s violation(s)\n' "$violations" >&2 - exit 1 -fi -echo "check-dashboard-service-names: $checked selector name(s) checked, clean" >&2 diff --git a/scripts/pre-push b/scripts/pre-push index 7f36a492c..f722350df 100755 --- a/scripts/pre-push +++ b/scripts/pre-push @@ -39,11 +39,5 @@ if ! nix shell --inputs-from . nixpkgs#jq --command sh "$repo_root/scripts/check exit 1 fi -echo "pre-push: running dashboard-service-name lint..." >&2 -if ! sh "$repo_root/scripts/check-dashboard-service-names.sh"; then - echo "pre-push: dashboard-service-name lint FAILED — fix before pushing" >&2 - exit 1 -fi - echo "pre-push: lints passed ✓" >&2 exit 0