Expose load test ID Prometheus gauge

CHANGELOG.md

@@ -1,7 +1,15 @@
 # Changelog
 
+## v0.8.0
+* [\#42](https://github.com/interchainio/tm-load-test/pull/42) - Add Prometheus
+  gauge for when load test is underway. This indicator exposes a customizable
+  load test ID.
+
+## v0.7.1
+* Re-released due to v0.7.0 being incorrectly tagged
+
 ## v0.7.0
-* [\#39](https://github.com/interchainio/tm-load-test/issues/39) - Add basic
+* [\#39](https://github.com/interchainio/tm-load-test/pull/40) - Add basic
   aggregate statistics output to CSV file.
 * Added integration test for standalone execution happy path.
 

README.md

@@ -146,6 +146,8 @@ The following kinds of metrics are made available here:
   * 5 = Slave completed load testing successfully
 * Standard Prometheus-provided metrics about the garbage collector in 
   `tm-load-test`
+* The ID of the load test currently underway (defaults to 0), set by way of the
+  `--load-test-id` flag on the master
 
 ## Aggregate Statistics
 As of `tm-load-test` v0.7.0, one can now write simple aggregate statistics to

pkg/loadtest/cli.go

@@ -12,7 +12,7 @@ import (
 )
 
 // CLIVersion must be manually updated as new versions are released.
-const CLIVersion = "v0.7.0"
+const CLIVersion = "v0.8.0"
 
 // cliVersionCommitID must be set through linker settings. See
 // https://stackoverflow.com/a/11355611/1156132 for details.
@@ -91,6 +91,7 @@ func buildCLI(cli *CLIConfig, logger logging.Logger) *cobra.Command {
 	masterCmd.PersistentFlags().IntVar(&masterCfg.ExpectSlaves, "expect-slaves", 2, "The number of slaves to expect to connect to the master before starting load testing")
 	masterCmd.PersistentFlags().IntVar(&masterCfg.SlaveConnectTimeout, "connect-timeout", 180, "The maximum number of seconds to wait for all slaves to connect")
 	masterCmd.PersistentFlags().IntVar(&masterCfg.ShutdownWait, "shutdown-wait", 0, "The number of seconds to wait after testing completes prior to shutting down the web server")
+	masterCmd.PersistentFlags().IntVar(&masterCfg.LoadTestID, "load-test-id", 0, "The ID of the load test currently underway")
 
 	var slaveCfg SlaveConfig
 	slaveCmd := &cobra.Command{

pkg/loadtest/config.go

@@ -44,6 +44,7 @@ type MasterConfig struct {
 	ExpectSlaves        int    `json:"expect_slaves"`   // The number of slaves to expect before starting the load test.
 	SlaveConnectTimeout int    `json:"connect_timeout"` // The number of seconds to wait for all slaves to connect.
 	ShutdownWait        int    `json:"shutdown_wait"`   // The number of seconds to wait at shutdown (while keeping the HTTP server running - primarily to allow Prometheus to keep polling).
+	LoadTestID          int    `json:"load_test_id"`    // An integer greater than 0 that will be exposed via a Prometheus gauge while the load test is underway.
 }
 
 // SlaveConfig is the configuration options specific to a slave node.
@@ -126,6 +127,9 @@ func (c MasterConfig) Validate() error {
 	if c.SlaveConnectTimeout < 1 {
 		return fmt.Errorf("master connect-timeout must be at least 1 second")
 	}
+	if c.LoadTestID < 0 {
+		return fmt.Errorf("master load-test-id must be 0 or greater")
+	}
 	return nil
 }
 

pkg/loadtest/master.go

@@ -60,6 +60,7 @@ type Master struct {
 	txRateMetric          prometheus.Gauge // The transaction throughput rate (tx/sec) as measured by the master since the last metrics update.
 	overallTxRateMetric   prometheus.Gauge // The overall transaction throughput rate (tx/sec) as measured by the master since the beginning of the load test.
 	slavesCompletedMetric prometheus.Gauge // The total number of slaves that have completed their testing.
+	testUnderwayMetric    prometheus.Gauge // The ID of the load test currently underway (-1 if none).
 
 	mtx       sync.Mutex
 	cancelled bool
@@ -113,6 +114,10 @@ func NewMaster(cfg *Config, masterCfg *MasterConfig) *Master {
 			Name: "tmloadtest_master_slaves_completed",
 			Help: "The total number of slaves that have completed their testing so far",
 		}),
+		testUnderwayMetric: promauto.NewGauge(prometheus.GaugeOpts{
+			Name: "tmloadtest_master_test_underway",
+			Help: "The ID of the load test currently underway (-1 if none)",
+		}),
 	}
 	mux := http.NewServeMux()
 	mux.HandleFunc("/", master.newWebSocketHandler())
@@ -123,6 +128,7 @@ func NewMaster(cfg *Config, masterCfg *MasterConfig) *Master {
 	}
 	master.svr = svr
 	master.stateMetric.Set(masterStarting)
+	master.testUnderwayMetric.Set(-1)
 	return master
 }
 
@@ -229,6 +235,12 @@ func (m *Master) receiveTestingUpdates() error {
 	m.logger.Info("Watching for slave updates")
 	m.stateMetric.Set(masterTesting)
 
+	// we set the current test underway ID to our configured load test ID for
+	// the duration of the test
+	m.testUnderwayMetric.Set(float64(m.masterCfg.LoadTestID))
+	// and we set it to -1 the moment all slaves are done
+	defer m.testUnderwayMetric.Set(-1)
+
 	completed := 0
 
 	progressTicker := time.NewTicker(masterProgressUpdateInterval)