feat(cmn): have BSV2 export metadata metrics (#2736)

Updates BaseServiceV2 to require a version string. Also now exports
parameters in a synthetic metric but ignores any parameters explicitly
marked secret.

.changeset/rude-eggs-pull.md

+---
+'@eth-optimism/common-ts': minor
+'@eth-optimism/drippie-mon': minor
+'@eth-optimism/fault-detector': minor
+'@eth-optimism/message-relayer': minor
+'@eth-optimism/replica-healthcheck': minor
+---
+
+BaseServiceV2 exposes service name and version as standard synthetic metric

packages/common-ts/src/base-service/base-service-v2.ts

@@ -13,7 +13,7 @@ import bodyParser from 'body-parser'
 import morgan from 'morgan'
 
 import { Logger } from '../common/logger'
-import { Metric } from './metrics'
+import { Metric, Gauge, Counter } from './metrics'
 import { validators } from './validators'
 
 export type Options = {
@@ -31,6 +31,7 @@ export type OptionsSpec<TOptions extends Options> = {
     validator: (spec?: Spec<TOptions[P]>) => ValidatorSpec<TOptions[P]>
     desc: string
     default?: TOptions[P]
+    secret?: boolean
   }
 }
 
@@ -38,6 +39,11 @@ export type MetricsV2 = {
   [key: string]: Metric
 }
 
+export type StandardMetrics = {
+  metadata: Gauge
+  unhandledErrors: Counter
+}
+
 export type MetricsSpec<TMetrics extends MetricsV2> = {
   [P in keyof Required<TMetrics>]: {
     type: new (configuration: any) => TMetrics[P]
@@ -99,7 +105,7 @@ export abstract class BaseServiceV2<
   /**
    * Metrics.
    */
-  protected readonly metrics: TMetrics
+  protected readonly metrics: TMetrics & StandardMetrics
 
   /**
    * Registry for prometheus metrics.
@@ -137,6 +143,7 @@ export abstract class BaseServiceV2<
    */
   constructor(params: {
     name: string
+    version: string
     optionsSpec: OptionsSpec<TOptions>
     metricsSpec: MetricsSpec<TMetrics>
     options?: Partial<TOptions>
@@ -170,6 +177,31 @@ export abstract class BaseServiceV2<
       },
     }
 
+    // List of options that can safely be logged.
+    const publicOptionNames = Object.entries(params.optionsSpec)
+      .filter(([, spec]) => {
+        return spec.secret !== true
+      })
+      .map(([key]) => {
+        return key
+      })
+
+    // Add default metrics to metrics spec.
+    ;(params.metricsSpec as any) = {
+      ...(params.metricsSpec || {}),
+
+      // Users cannot set these options.
+      metadata: {
+        type: Gauge,
+        desc: 'Service metadata',
+        labels: ['name', 'version'].concat(publicOptionNames),
+      },
+      unhandledErrors: {
+        type: Counter,
+        desc: 'Unhandled errors',
+      },
+    }
+
     /**
      * Special snake_case function which accounts for the common strings "L1" and "L2" which would
      * normally be split into "L_1" and "L_2" by the snake_case function.
@@ -274,7 +306,7 @@ export abstract class BaseServiceV2<
         labelNames: spec.labels || [],
       })
       return acc
-    }, {}) as TMetrics
+    }, {}) as TMetrics & StandardMetrics
 
     // Create the metrics server.
     this.metricsRegistry = prometheus.register
@@ -309,6 +341,19 @@ export abstract class BaseServiceV2<
     // Handle stop signals.
     process.on('SIGTERM', stop)
     process.on('SIGINT', stop)
+
+    // Set metadata synthetic metric.
+    this.metrics.metadata.set(
+      {
+        name: params.name,
+        version: params.version,
+        ...publicOptionNames.reduce((acc, key) => {
+          acc[key] = config.str(key)
+          return acc
+        }, {}),
+      },
+      1
+    )
   }
 
   /**
@@ -330,21 +375,7 @@ export abstract class BaseServiceV2<
       app.use(bodyParser.urlencoded({ extended: true }))
 
       // Logging.
-      app.use(
-        morgan((tokens, req, res) => {
-          return [
-            tokens.method(req, res),
-            tokens.url(req, res),
-            tokens.status(req, res),
-            JSON.stringify(req.body),
-            '\n',
-            tokens.res(req, res, 'content-length'),
-            '-',
-            tokens['response-time'](req, res),
-            'ms',
-          ].join(' ')
-        })
-      )
+      app.use(morgan('short'))
 
       // Metrics.
       // Will expose a /metrics endpoint by default.
@@ -397,6 +428,7 @@ export abstract class BaseServiceV2<
         try {
           await this.main()
         } catch (err) {
+          this.metrics.unhandledErrors.inc()
           this.logger.error('caught an unhandled exception', {
             message: err.message,
             stack: err.stack,

packages/drippie-mon/src/service.ts

@@ -14,7 +14,6 @@ type DrippieMonOptions = {
 }
 
 type DrippieMonMetrics = {
-  metadata: Gauge
   isExecutable: Gauge
   executedDripCount: Gauge
   unexpectedRpcErrors: Counter
@@ -31,6 +30,8 @@ export class DrippieMonService extends BaseServiceV2<
 > {
   constructor(options?: Partial<DrippieMonOptions>) {
     super({
+      // eslint-disable-next-line @typescript-eslint/no-var-requires
+      version: require('../package.json').version,
       name: 'drippie-mon',
       loop: true,
       loopIntervalMs: 60_000,
@@ -39,6 +40,7 @@ export class DrippieMonService extends BaseServiceV2<
         rpc: {
           validator: validators.provider,
           desc: 'Provider for network where Drippie is deployed',
+          secret: true,
         },
         drippieAddress: {
           validator: validators.str,
@@ -46,11 +48,6 @@ export class DrippieMonService extends BaseServiceV2<
         },
       },
       metricsSpec: {
-        metadata: {
-          type: Gauge,
-          desc: 'Drippie Monitor metadata',
-          labels: ['version', 'address'],
-        },
         isExecutable: {
           type: Gauge,
           desc: 'Whether or not the drip is currently executable',
@@ -76,15 +73,6 @@ export class DrippieMonService extends BaseServiceV2<
       DrippieArtifact.abi,
       this.options.rpc
     )
-
-    this.metrics.metadata.set(
-      {
-        // eslint-disable-next-line @typescript-eslint/no-var-requires
-        version: require('../package.json').version,
-        address: this.options.drippieAddress,
-      },
-      1
-    )
   }
 
   protected async main(): Promise<void> {

packages/fault-detector/src/service.ts

@@ -32,6 +32,8 @@ type State = {
 export class FaultDetector extends BaseServiceV2<Options, Metrics, State> {
   constructor(options?: Partial<Options>) {
     super({
+      // eslint-disable-next-line @typescript-eslint/no-var-requires
+      version: require('../package.json').version,
       name: 'fault-detector',
       loop: true,
       loopIntervalMs: 1000,
@@ -40,10 +42,12 @@ export class FaultDetector extends BaseServiceV2<Options, Metrics, State> {
         l1RpcProvider: {
           validator: validators.provider,
           desc: 'Provider for interacting with L1',
+          secret: true,
         },
         l2RpcProvider: {
           validator: validators.provider,
           desc: 'Provider for interacting with L2',
+          secret: true,
         },
         startBatchIndex: {
           validator: validators.num,

packages/message-relayer/src/service.ts

@@ -37,20 +37,25 @@ export class MessageRelayerService extends BaseServiceV2<
 > {
   constructor(options?: Partial<MessageRelayerOptions>) {
     super({
-      name: 'Message_Relayer',
+      // eslint-disable-next-line @typescript-eslint/no-var-requires
+      version: require('../package.json').version,
+      name: 'message-relayer',
       options,
       optionsSpec: {
         l1RpcProvider: {
           validator: validators.provider,
           desc: 'Provider for interacting with L1.',
+          secret: true,
         },
         l2RpcProvider: {
           validator: validators.provider,
           desc: 'Provider for interacting with L2.',
+          secret: true,
         },
         l1Wallet: {
           validator: validators.wallet,
           desc: 'Wallet used to interact with L1.',
+          secret: true,
         },
         fromL2TransactionIndex: {
           validator: validators.num,

packages/replica-healthcheck/src/service.ts

@@ -32,17 +32,21 @@ export class HealthcheckService extends BaseServiceV2<
 > {
   constructor(options?: Partial<HealthcheckOptions>) {
     super({
-      name: 'Healthcheck',
+      // eslint-disable-next-line @typescript-eslint/no-var-requires
+      version: require('../package.json').version,
+      name: 'healthcheck',
       loopIntervalMs: 5000,
       options,
       optionsSpec: {
         referenceRpcProvider: {
           validator: validators.provider,
           desc: 'Provider for interacting with L1',
+          secret: true,
         },
         targetRpcProvider: {
           validator: validators.provider,
           desc: 'Provider for interacting with L2',
+          secret: true,
         },
         onDivergenceWaitMs: {
           validator: validators.num,