feat: improve gateway.yaml path handling and Olric client initialization

- Enhanced the DefaultPath function to remember the preferred data path for gateway.yaml, allowing for better error messaging and fallback options.
- Introduced a new function to initialize the Olric client with retry logic, improving resilience during client setup and providing clearer logging for connection attempts.
- Updated logging to provide detailed feedback on Olric client initialization attempts, enhancing troubleshooting capabilities.

CHANGELOG.md

@@ -13,6 +13,21 @@ The format is based on [Keep a Changelog][keepachangelog] and adheres to [Semant
 ### Deprecated
 
 ### Fixed
+## [0.69.13] - 2025-11-14
+
+### Added
+\n
+### Changed
+- The Gateway service now waits for the Olric cache service to start before attempting initialization.
+- Improved robustness of Olric cache client initialization with retry logic and exponential backoff.
+
+### Deprecated
+
+### Removed
+
+### Fixed
+- Corrected the default path logic for 'gateway.yaml' to prioritize the production data directory while maintaining fallback to legacy paths.
+
 ## [0.69.12] - 2025-11-14
 
 ### Added

Makefile

@@ -19,7 +19,7 @@ test-e2e:
 
 .PHONY: build clean test run-node run-node2 run-node3 run-example deps tidy fmt vet lint clear-ports install-hooks kill
 
-VERSION := 0.69.12
+VERSION := 0.69.13
 COMMIT  ?= $(shell git rev-parse --short HEAD 2>/dev/null || echo unknown)
 DATE    ?= $(shell date -u +%Y-%m-%dT%H:%M:%SZ)
 LDFLAGS := -X 'main.version=$(VERSION)' -X 'main.commit=$(COMMIT)' -X 'main.date=$(DATE)'

pkg/config/paths.go

@@ -42,14 +42,15 @@ func DefaultPath(component string) (string, error) {
 		return "", err
 	}
 
+	var gatewayDefault string
 	// For gateway.yaml, check data/ directory first (production location)
 	if component == "gateway.yaml" {
 		dataPath := filepath.Join(dir, "data", component)
 		if _, err := os.Stat(dataPath); err == nil {
 			return dataPath, nil
 		}
-		// Return data path as default for gateway.yaml (even if it doesn't exist yet)
+		// Remember the preferred default so we can still fall back to legacy paths
-		return dataPath, nil
+		gatewayDefault = dataPath
 	}
 
 	// First check in ~/.debros/configs/ (production installer location)
@@ -64,6 +65,12 @@ func DefaultPath(component string) (string, error) {
 		return legacyPath, nil
 	}
 
+	if gatewayDefault != "" {
+		// If we preferred the data path (gateway.yaml) but didn't find it anywhere else,
+		// return the data path so error messages point to the production location.
+		return gatewayDefault, nil
+	}
+
 	// Return configs path as default (even if it doesn't exist yet)
 	// This allows the error message to show the expected production location
 	return configsPath, nil

pkg/environments/production/services.go

@@ -233,11 +233,12 @@ WantedBy=multi-user.target
 // GenerateGatewayService generates the DeBros Gateway systemd unit
 func (ssg *SystemdServiceGenerator) GenerateGatewayService(nodeType string) string {
 	nodeService := fmt.Sprintf("debros-node-%s.service", nodeType)
+	olricService := "debros-olric.service"
 	logFile := filepath.Join(ssg.debrosDir, "logs", "gateway.log")
 	return fmt.Sprintf(`[Unit]
 Description=DeBros Gateway
-After=%s
+After=%s %s
-Wants=%s
+Wants=%s %s
 
 [Service]
 Type=simple
@@ -262,7 +263,7 @@ ReadWritePaths=%s
 
 [Install]
 WantedBy=multi-user.target
-`, nodeService, nodeService, ssg.debrosHome, ssg.debrosHome, ssg.debrosHome, ssg.debrosDir, logFile, logFile, ssg.debrosDir)
+`, nodeService, olricService, nodeService, olricService, ssg.debrosHome, ssg.debrosHome, ssg.debrosHome, ssg.debrosDir, logFile, logFile, ssg.debrosDir)
 }
 
 // SystemdController manages systemd service operations

pkg/gateway/gateway.go

@@ -5,6 +5,7 @@ import (
 	"crypto/rand"
 	"crypto/rsa"
 	"database/sql"
+	"fmt"
 	"net"
 	"os"
 	"path/filepath"
@@ -25,6 +26,12 @@ import (
 	_ "github.com/rqlite/gorqlite/stdlib"
 )
 
+const (
+	olricInitMaxAttempts    = 5
+	olricInitInitialBackoff = 500 * time.Millisecond
+	olricInitMaxBackoff     = 5 * time.Second
+)
+
 // Config holds configuration for the gateway server
 type Config struct {
 	ListenAddr      string
@@ -182,7 +189,7 @@ func New(logger *logging.ColoredLogger, cfg *Config) (*Gateway, error) {
 		Servers: olricServers,
 		Timeout: cfg.OlricTimeout,
 	}
-	olricClient, olricErr := olric.NewClient(olricCfg, logger.Logger)
+	olricClient, olricErr := initializeOlricClientWithRetry(olricCfg, logger)
 	if olricErr != nil {
 		logger.ComponentWarn(logging.ComponentGeneral, "failed to initialize Olric cache client; cache endpoints disabled", zap.Error(olricErr))
 	} else {
@@ -330,6 +337,38 @@ func (g *Gateway) getLocalSubscribers(topic, namespace string) []*localSubscribe
 	return nil
 }
 
+func initializeOlricClientWithRetry(cfg olric.Config, logger *logging.ColoredLogger) (*olric.Client, error) {
+	backoff := olricInitInitialBackoff
+
+	for attempt := 1; attempt <= olricInitMaxAttempts; attempt++ {
+		client, err := olric.NewClient(cfg, logger.Logger)
+		if err == nil {
+			if attempt > 1 {
+				logger.ComponentInfo(logging.ComponentGeneral, "Olric cache client initialized after retries",
+					zap.Int("attempts", attempt))
+			}
+			return client, nil
+		}
+
+		logger.ComponentWarn(logging.ComponentGeneral, "Olric cache client init attempt failed",
+			zap.Int("attempt", attempt),
+			zap.Duration("retry_in", backoff),
+			zap.Error(err))
+
+		if attempt == olricInitMaxAttempts {
+			return nil, fmt.Errorf("failed to initialize Olric cache client after %d attempts: %w", attempt, err)
+		}
+
+		time.Sleep(backoff)
+		backoff *= 2
+		if backoff > olricInitMaxBackoff {
+			backoff = olricInitMaxBackoff
+		}
+	}
+
+	return nil, fmt.Errorf("failed to initialize Olric cache client")
+}
+
 // discoverOlricServers discovers Olric server addresses from LibP2P peers
 // Returns a list of IP:port addresses where Olric servers are expected to run (port 3320)
 func discoverOlricServers(networkClient client.NetworkClient, logger *zap.Logger) []string {