Make MaxFederationRetries configurable

federationsender/federationsender.go

@@ -49,8 +49,10 @@ func NewInternalAPI(
 	}
 
 	stats := &statistics.Statistics{
-		DB: federationSenderDB,
+		DB:                     federationSenderDB,
+		FailuresUntilBlacklist: base.Cfg.Matrix.FederationMaxRetries,
 	}
+
 	queues := queue.NewOutgoingQueues(
 		federationSenderDB, base.Cfg.Matrix.ServerName, federation, rsAPI, stats,
 		&queue.SigningInfo{

federationsender/statistics/statistics.go

@@ -11,13 +11,6 @@ import (
 	"go.uber.org/atomic"
 )
 
-const (
-	// How many times should we tolerate consecutive failures before we
-	// just blacklist the host altogether? Bear in mind that the backoff
-	// is exponential, so the max time here to attempt is 2**failures.
-	FailuresUntilBlacklist = 3 // 16 equates to roughly 18 hours.
-)
-
 // Statistics contains information about all of the remote federated
 // hosts that we have interacted with. It is basically a threadsafe
 // wrapper.
@@ -25,6 +18,11 @@ type Statistics struct {
 	DB      storage.Database
 	servers map[gomatrixserverlib.ServerName]*ServerStatistics
 	mutex   sync.RWMutex
+
+	// How many times should we tolerate consecutive failures before we
+	// just blacklist the host altogether? Bear in mind that the backoff
+	// is exponential, so the max time here to attempt is 2**failures.
+	FailuresUntilBlacklist uint32
 }
 
 // ForServer returns server statistics for the given server name. If it
@@ -95,7 +93,7 @@ func (s *ServerStatistics) Failure() bool {
 	failCounter := s.failCounter.Add(1)
 
 	// Check that we haven't failed more times than is acceptable.
-	if failCounter >= FailuresUntilBlacklist {
+	if failCounter >= s.statistics.FailuresUntilBlacklist {
 		// We've exceeded the maximum amount of times we're willing
 		// to back off, which is probably in the region of hours by
 		// now. Mark the host as blacklisted and tell the caller to

internal/config/config.go

@@ -102,6 +102,11 @@ type Dendrite struct {
 		// Perspective keyservers, to use as a backup when direct key fetch
 		// requests don't succeed
 		KeyPerspectives KeyPerspectives `yaml:"key_perspectives"`
+		// Federation failure threshold. How many consecutive failures that we should
+		// tolerate when sending federation requests to a specific server. The backoff
+		// is 2**x seconds, so 1 = 2 seconds, 2 = 4 seconds, 3 = 8 seconds, etc.
+		// The default value is 16 if not specified, which is circa 18 hours.
+		FederationMaxRetries uint32 `yaml:"federation_max_retries"`
 	} `yaml:"matrix"`
 
 	// The configuration specific to the media repostitory.
@@ -479,6 +484,10 @@ func (config *Dendrite) SetDefaults() {
 		config.Matrix.TrustedIDServers = []string{}
 	}
 
+	if config.Matrix.FederationMaxRetries == 0 {
+		config.Matrix.FederationMaxRetries = 16
+	}
+
 	if config.Media.MaxThumbnailGenerators == 0 {
 		config.Media.MaxThumbnailGenerators = 10
 	}