config/console: make cluster operation timeouts configurable

Author: alessandro-grassi

backend/pkg/config/console.go

@@ -22,12 +22,14 @@ const DefaultMaxDeserializationPayloadSize = 20_480 // 20 KB
 type Console struct {
 	TopicDocumentation ConsoleTopicDocumentation `yaml:"topicDocumentation"`
 	API                ConsoleAPI                `yaml:"api"`
+	OperationTimeouts  OperationTimeouts         `yaml:"operationTimeouts"`
 }
 
 // SetDefaults for Console configs.
 func (c *Console) SetDefaults() {
 	c.TopicDocumentation.SetDefaults()
 	c.API.SetDefaults()
+	c.OperationTimeouts.SetDefaults()
 }
 
 // RegisterFlags for sensitive Console configurations.
@@ -45,5 +47,9 @@ func (c *Console) Validate() error {
 		return fmt.Errorf("failed to validate API config: %w", err)
 	}
 
+	if err := c.OperationTimeouts.Validate(); err != nil {
+		return fmt.Errorf("failed to validate timeouts config: %w", err)
+	}
+
 	return nil
 }

backend/pkg/config/operation_timeouts.go

@@ -0,0 +1,48 @@
+// Copyright 2024 Redpanda Data, Inc.
+//
+// Use of this software is governed by the Business Source License
+// included in the file licenses/BSL.md
+//
+// As of the Change Date specified in that file, in accordance with
+// the Business Source License, use of this software will be governed
+// by the Apache License, Version 2.0
+
+package config
+
+import (
+	"errors"
+	"time"
+)
+
+// OperationTimeouts contains timeout configurations for various Kafka operations
+// performed by the console service.
+type OperationTimeouts struct {
+	// ClusterInfo is the timeout applied when fetching broker metadata and log dir
+	// information to build the cluster overview. A shorter timeout avoids long
+	// response times when a single broker is temporarily unreachable, but may
+	// cause incomplete results on slow clusters.
+	ClusterInfo time.Duration `yaml:"clusterInfo"`
+
+	// TopicsOverview is the timeout applied when fetching topic metadata and per-topic
+	// log dir sizes to build the topics list. A shorter timeout avoids long response
+	// times when a single broker is temporarily unreachable, but may cause incomplete
+	// results on slow clusters.
+	TopicsOverview time.Duration `yaml:"topicsOverview"`
+}
+
+// SetDefaults for ConsoleTimeouts.
+func (c *OperationTimeouts) SetDefaults() {
+	c.ClusterInfo = 6 * time.Second
+	c.TopicsOverview = 5 * time.Second
+}
+
+// Validate ConsoleTimeouts.
+func (c *OperationTimeouts) Validate() error {
+	if c.ClusterInfo <= 0 {
+		return errors.New("console.operationTimeouts.clusterInfo must be a positive duration")
+	}
+	if c.TopicsOverview <= 0 {
+		return errors.New("console.operationTimeouts.topicsOverview must be a positive duration")
+	}
+	return nil
+}

backend/pkg/console/cluster_info.go

@@ -13,7 +13,6 @@ import (
 	"context"
 	"log/slog"
 	"sort"
-	"time"
 
 	"github.com/twmb/franz-go/pkg/kmsg"
 	"golang.org/x/sync/errgroup"
@@ -53,7 +52,7 @@ func (s *Service) GetClusterInfo(ctx context.Context) (*ClusterInfo, error) {
 
 	// We use a child context with a shorter timeout because otherwise we'll potentially have very long response
 	// times in case of a single broker being down.
-	childCtx, cancel := context.WithTimeout(egCtx, 6*time.Second)
+	childCtx, cancel := context.WithTimeout(egCtx, s.cfg.Console.OperationTimeouts.ClusterInfo)
 	defer cancel()
 
 	eg.Go(func() error {

backend/pkg/console/topic_overview.go

@@ -15,7 +15,6 @@ import (
 	"log/slog"
 	"sort"
 	"sync"
-	"time"
 )
 
 // DocumentationState denotes whether topic documentation is available for a certain
@@ -77,7 +76,7 @@ func (s *Service) GetTopicsOverview(ctx context.Context) ([]*TopicSummary, error
 
 	// 3. Get log dir sizes & configs for each topic concurrently
 	// Use a shorter ctx timeout so that we don't wait for too long if one broker is currently down.
-	childCtx, cancel := context.WithTimeout(ctx, 5*time.Second)
+	childCtx, cancel := context.WithTimeout(ctx, s.cfg.Console.OperationTimeouts.TopicsOverview)
 	defer cancel()
 
 	configs := make(map[string]*TopicConfig)