Enhancement: Reload all scrape pools concurrently (#16595)

* Reload all scrape pools concurrently

At the moment all scrape pools that need to be reloaded are reloaded one by one. While reloads are ongoing mtxScrape is locked.
For each pool that's being reloaded we need to wait until all targets are updated.
This whole process can take a while and the more scrape pools to reload the longer.
At the same time all pools are independent and there's no real reason to do them one-by-one.
Reload each pool in a seperate goroutine so we finish config reload as ASAP as possible and unlock the mtxScrape.

Signed-off-by: Lukasz Mierzwa <l.mierzwa@gmail.com>

* Address PR review feedback

Signed-off-by: Lukasz Mierzwa <l.mierzwa@gmail.com>

---------

Signed-off-by: Lukasz Mierzwa <l.mierzwa@gmail.com>

scrape/manager.go

@@ -26,6 +26,7 @@ import (
 	config_util "github.com/prometheus/common/config"
 	"github.com/prometheus/common/model"
 	"github.com/prometheus/common/promslog"
+	"go.uber.org/atomic"
 
 	"github.com/prometheus/prometheus/config"
 	"github.com/prometheus/prometheus/discovery/targetgroup"
@@ -287,17 +288,27 @@ func (m *Manager) ApplyConfig(cfg *config.Config) error {
 	}
 
 	// Cleanup and reload pool if the configuration has changed.
-	var failed bool
+	var (
-	for name, sp := range m.scrapePools {
+		failed   atomic.Bool
-		switch cfg, ok := m.scrapeConfigs[name]; {
+		wg       sync.WaitGroup
+		toDelete sync.Map // Stores the list of names of pools to delete.
+	)
+	for poolName, pool := range m.scrapePools {
+		wg.Add(1)
+		cfg, ok := m.scrapeConfigs[poolName]
+		// Reload each scrape pool in a dedicated goroutine so we don't have to wait a long time
+		// if we have a lot of scrape pools to update.
+		go func(name string, sp *scrapePool, cfg *config.ScrapeConfig, ok bool) {
+			defer wg.Done()
+			switch {
 			case !ok:
 				sp.stop()
-			delete(m.scrapePools, name)
+				toDelete.Store(name, struct{}{})
 			case !reflect.DeepEqual(sp.config, cfg):
 				err := sp.reload(cfg)
 				if err != nil {
 					m.logger.Error("error reloading scrape pool", "err", err, "scrape_pool", name)
-				failed = true
+					failed.Store(true)
 				}
 				fallthrough
 			case ok:
@@ -307,9 +318,16 @@ func (m *Manager) ApplyConfig(cfg *config.Config) error {
 					sp.logger.Error("No logger found. This is a bug in Prometheus that should be reported upstream.", "scrape_pool", name)
 				}
 			}
+		}(poolName, pool, cfg, ok)
 	}
+	wg.Wait()
 
-	if failed {
+	toDelete.Range(func(name, _ any) bool {
+		delete(m.scrapePools, name.(string))
+		return true
+	})
+
+	if failed.Load() {
 		return errors.New("failed to apply the new configuration")
 	}
 	return nil