[*] deprecate --direct-os-stats flag

cmd/pgwatch/doc.go

@@ -27,18 +27,13 @@
 //	                                         all monitored DBs when missing. Main
 //	                                         usage - pg_stat_statements
 //	                                         [$PW_TRY_CREATE_LISTED_EXTS_IF_MISSING]
-//
+//	    --create-helpers                     Create helper database objects from
+//	                                         metric definitions
+//	                                         [$PW_CREATE_HELPERS]
 // Metrics:
 //
 //	-m, --metrics=                           File or folder of YAML files with
 //	                                         metrics definitions [$PW_METRICS]
-//	    --create-helpers                     Create helper database objects from
-//	                                         metric definitions
-//	                                         [$PW_CREATE_HELPERS]
-//	    --direct-os-stats                    Extract OS related psutil statistics
-//	                                         not via PL/Python wrappers but
-//	                                         directly on host
-//	                                         [$PW_DIRECT_OS_STATS]
 //	    --instance-level-cache-max-seconds=  Max allowed staleness for instance
 //	                                         level metric data shared between DBs
 //	                                         of an instance. Affects 'continuous'

cmd/pgwatch/main.go

@@ -98,6 +98,10 @@ func main() {
 		return
 	}
 
+	if opts.Metrics.DirectOSStats {
+		logger.Warning("--direct-os-stats flag is deprecated, direct OS access is now applied automatically for relevant metrics if on same host.")
+	}
+
 	reaper := reaper.NewReaper(mainCtx, opts)
 
 	if _, err = webserver.Init(mainCtx, opts.WebUI, webui.WebUIFs, opts.MetricsReaderWriter,

cmd/pgwatch/main_integration_test.go

@@ -50,6 +50,9 @@ metrics:
     test_metric:
         sqls:
             11:  select (extract(epoch from now()) * 1e9)::int8 as epoch_ns, 42 as test_metric
+    cpu_load:
+        sqls:
+            11:  select 'metric fetched via PostgreSQL not OS'
 `), 0644))
 
 	pg, err := initTestContainer()
@@ -66,6 +69,7 @@ metrics:
   is_enabled: true
   custom_metrics:
     test_metric: 60
+    cpu_load: 60
 `), 0644))
 
 	// Mock Exit to capture exit code
@@ -127,4 +131,24 @@ metrics:
 		<-mainCtx.Done() // Wait for main to finish
 		assert.Equal(t, cmdopts.ExitCodeWebUIError, gotExit, "port should be busy and fail to bind")
 	})
+
+	t.Run("non-direct os stats", func(t *testing.T) {
+		os.Remove(jsonSink) // truncate output file
+		os.Args = []string{
+			"pgwatch",
+			"--sources", sourcesYaml,
+			"--metrics", metricsYaml,
+			"--sink", "jsonfile://" + jsonSink,
+			"--web-disable",
+		}
+
+		go main()
+		<-time.After(5 * time.Second) // Wait for main to fetch metric
+		cancel()
+		<-mainCtx.Done() // Wait for main to finish
+
+		datab, err := os.ReadFile(jsonSink)
+		assert.NoError(t, err)
+		assert.Contains(t, string(datab), "metric fetched via PostgreSQL not OS")
+	})
 }

docs/reference/cli_env.md

@@ -63,11 +63,6 @@ It reads the configuration from the specified sources and metrics, then begins c
     Postgres URI, YAML file or folder of YAML files with metrics definitions.  
     ENV: `$PW_METRICS`
 
-- `--direct-os-stats`
-
-    Extract OS related psutil statistics not via PL/Python wrappers but directly on host.  
-    ENV: `$PW_DIRECT_OS_STATS`
-
 - `--instance-level-cache-max-seconds=`
 
     Max allowed staleness for instance level metric data shared between DBs of an instance. Set to 0 to disable (default: 30).  

docs/tutorial/preparing_databases.md

@@ -190,21 +190,21 @@ pgwatch metric print-init psutil_cpu psutil_mem psutil_disk psutil_disk_io_total
     based on the Linux distro / Kernel version used, so small
     adjustments might be needed there (e.g. to remove a non-existent
     column). Minimum usable Kernel version required is 3.3.
-- When running the gatherer locally one can enable the `--direct-os-stats`
-    parameter to signal that we can fetch the data for the default `psutil*` metrics
-    directly from OS counters. If direct OS fetching fails though, the
-    fallback is still to try via PL/Python wrappers.
+- When pgwatch runs on the same host as a monitored source,
+    it auto-detects this and tries to fetch the default `psutil*`
+    metrics data directly from OS counters. If this direct OS fetch fails,
+    it falls back to using PL/Python wrappers.
 - In rare cases when some "helpers" have been installed, and when
     doing a binary PostgreSQL upgrade at some later point in time via
     `pg_upgrade`, this could result in error messages
     thrown. Then just drop those failing helpers on the "to be
     upgraded" cluster and re-create them after the upgrade process.
 
 !!! Info
-    If despite all the warnings you still want to run the pgwatch
-    with a sufficient user account (e.g. a superuser) you can also
+    If, despite all the warnings, you still want to run pgwatch
+    with a sufficient user account (e.g., a superuser), you can
     use the `--create-helpers` parameter to automatically create all
-    needed helper functions in the monitored databases.
+    needed helper functions in the monitored databases on startup.
 
 ## Different source types explained
 

internal/db/conn.go

@@ -71,14 +71,16 @@ func MarshallParamToJSONB(v any) any {
 func IsClientOnSameHost(conn PgxIface) (bool, error) {
 	ctx := context.Background()
 
-	// Step 1: Check connection type using SQL
+	// Option 1: Check if connection type is unix socket
 	var isUnixSocket bool
 	err := conn.QueryRow(ctx, "SELECT COALESCE(inet_client_addr(), inet_server_addr()) IS NULL").Scan(&isUnixSocket)
 	if err != nil || isUnixSocket {
 		return isUnixSocket, err
 	}
 
-	// Step 2: Retrieve unique cluster identifier
+	// Option 2: Compare system identifier from file system vs cluster identifier
+
+	// Step 1: Retrieve unique cluster identifier
 	var dataDirectory string
 	if err := conn.QueryRow(ctx, "SHOW data_directory").Scan(&dataDirectory); err != nil {
 		return false, err
@@ -89,7 +91,7 @@ func IsClientOnSameHost(conn PgxIface) (bool, error) {
 		return false, err
 	}
 
-	// Step 3: Compare system identifier from file system
+	// Step 2: Retrieve system identifier from file system
 	pgControlFile := filepath.Join(dataDirectory, "global", "pg_control")
 	file, err := os.Open(pgControlFile)
 	if err != nil {
@@ -102,5 +104,6 @@ func IsClientOnSameHost(conn PgxIface) (bool, error) {
 		return false, err
 	}
 
+	// Compare file system identifier and cluster identifier
 	return fileSystemIdentifier == systemIdentifier, nil
 }

internal/metrics/cmdopts.go

@@ -7,7 +7,7 @@ import (
 // CmdOpts specifies metric command-line options
 type CmdOpts struct {
 	Metrics                      string `short:"m" long:"metrics" mapstructure:"metrics" description:"Postgres URI or path to YAML file with metrics definitions" env:"PW_METRICS"`
-	DirectOSStats                bool   `long:"direct-os-stats" mapstructure:"direct-os-stats" description:"Extract OS related psutil statistics not via PL/Python wrappers but directly on host" env:"PW_DIRECT_OS_STATS"`
+	DirectOSStats                bool   `hidden:"true" long:"direct-os-stats" mapstructure:"direct-os-stats" description:"Extract OS related psutil statistics not via PL/Python wrappers but directly on host" env:"PW_DIRECT_OS_STATS"`
 	InstanceLevelCacheMaxSeconds int64  `long:"instance-level-cache-max-seconds" mapstructure:"instance-level-cache-max-seconds" description:"Max allowed staleness for instance level metric data shared between DBs of an instance. Set to 0 to disable" env:"PW_INSTANCE_LEVEL_CACHE_MAX_SECONDS" default:"30"`
 	EmergencyPauseTriggerfile    string `long:"emergency-pause-triggerfile" mapstructure:"emergency-pause-triggerfile" description:"When the file exists no metrics will be temporarily fetched" env:"PW_EMERGENCY_PAUSE_TRIGGERFILE" default:"/tmp/pgwatch-emergency-pause"`
 }

internal/reaper/file.go

@@ -37,8 +37,8 @@ const (
 
 var directlyFetchableOSMetrics = []string{metricPsutilCPU, metricPsutilDisk, metricPsutilDiskIoTotal, metricPsutilMem, metricCPULoad}
 
-func IsDirectlyFetchableMetric(metric string) bool {
-	return slices.Contains(directlyFetchableOSMetrics, metric)
+func IsDirectlyFetchableMetric(md *sources.SourceConn, metric string) bool {
+	return slices.Contains(directlyFetchableOSMetrics, metric) && md.IsClientOnSameHost()
 }
 
 func (r *Reaper) FetchStatsDirectlyFromOS(ctx context.Context, md *sources.SourceConn, metricName string) (*metrics.MeasurementEnvelope, error) {

internal/reaper/metric_test.go

@@ -29,10 +29,13 @@ var (
 func TestReaper_FetchStatsDirectlyFromOS(t *testing.T) {
 	a := assert.New(t)
 	r := &Reaper{}
-	conn, _ := pgxmock.NewPool()
+	conn, _ := pgxmock.NewPool(pgxmock.QueryMatcherOption(pgxmock.QueryMatcherEqual))
+	expq := conn.ExpectQuery("SELECT COALESCE(inet_client_addr(), inet_server_addr()) IS NULL")
+	expq.Times(uint(len(directlyFetchableOSMetrics)))
 	md := &sources.SourceConn{Conn: conn}
 	for _, m := range directlyFetchableOSMetrics {
-		a.True(IsDirectlyFetchableMetric(m), "Expected %s to be directly fetchable", m)
+		expq.WillReturnRows(pgxmock.NewRows([]string{"is_unix_socket"}).AddRow(true))
+		a.True(IsDirectlyFetchableMetric(md, m), "Expected %s to be directly fetchable", m)
 		a.NotPanics(func() {
 			_, _ = r.FetchStatsDirectlyFromOS(context.Background(), md, m)
 		})

internal/reaper/reaper.go

@@ -312,14 +312,13 @@ func (r *Reaper) reapMetricMeasurements(ctx context.Context, md *sources.SourceC
 
 		var metricStoreMessages *metrics.MeasurementEnvelope
 
-		// 1st try local overrides for some metrics if operating in push mode
-		if r.Metrics.DirectOSStats && IsDirectlyFetchableMetric(metricName) {
-			metricStoreMessages, err = r.FetchStatsDirectlyFromOS(ctx, md, metricName)
-			if err != nil {
+		t1 := time.Now()
+		// 1st try local overrides for system metrics if operating on the same host
+		if IsDirectlyFetchableMetric(md, metricName) {
+			if metricStoreMessages, err = r.FetchStatsDirectlyFromOS(ctx, md, metricName); err != nil {
 				l.WithError(err).Errorf("Could not read metric directly from OS")
 			}
 		}
-		t1 := time.Now()
 		if metricStoreMessages == nil {
 			metricStoreMessages, err = r.FetchMetric(ctx, md, metricName)
 		}

internal/sources/conn.go

@@ -133,6 +133,12 @@ func (md *SourceConn) GetMetricInterval(name string) float64 {
 	return md.Metrics[name]
 }
 
+// IsClientOnSameHost checks if the pgwatch client is running on the same host as the PostgreSQL server
+func (md *SourceConn) IsClientOnSameHost() bool {
+	ok, err := db.IsClientOnSameHost(md.Conn)
+	return ok && err == nil
+}
+
 // SetDatabaseName sets the database name in the connection config for resolved databases
 func (md *SourceConn) SetDatabaseName(name string) {
 	if err := md.ParseConfig(); err != nil {