Fix Spotify 429 rate limiting and startup performance issues

- Fix: Use correct HttpClient (_webApiClient) for GraphQL library playlists endpoint
  - Was using _httpClient which pointed to wrong base URL causing 429 errors
- Add: Retry logic with Retry-After header support for 429 responses
- Add: Minimum 500ms delay between library playlist pages to prevent rate limiting
- Add: 5-second timeout per endpoint benchmark ping to prevent slow endpoints from blocking startup
- Add: Documentation for timeout requirements in EndpointBenchmarkService
- Fix: ARM64 compatibility for spotify-lyrics service via platform emulation in docker-compose

allstarr/Services/Common/EndpointBenchmarkService.cs

@@ -20,6 +20,9 @@ public class EndpointBenchmarkService
     /// <summary>
     /// Benchmarks a list of endpoints by making test requests.
     /// Returns endpoints sorted by average response time (fastest first).
+    /// 
+    /// IMPORTANT: The testFunc should implement its own timeout to prevent slow endpoints
+    /// from blocking startup. Recommended: 5-10 second timeout per ping.
     /// </summary>
     public async Task<List<string>> BenchmarkEndpointsAsync(
         List<string> endpoints, 

allstarr/Services/Spotify/SpotifyApiClient.cs

@@ -349,6 +349,17 @@ public class SpotifyApiClient : IDisposable
             
             var response = await _webApiClient.SendAsync(request, cancellationToken);
             
+            // Handle 429 rate limiting with exponential backoff
+            if (response.StatusCode == System.Net.HttpStatusCode.TooManyRequests)
+            {
+                var retryAfter = response.Headers.RetryAfter?.Delta ?? TimeSpan.FromSeconds(5);
+                _logger.LogWarning("Spotify rate limit hit (429) when fetching playlist {PlaylistId}. Waiting {Seconds}s before retry...", playlistId, retryAfter.TotalSeconds);
+                await Task.Delay(retryAfter, cancellationToken);
+                
+                // Retry the request
+                response = await _webApiClient.SendAsync(request, cancellationToken);
+            }
+            
             if (!response.IsSuccessStatusCode)
             {
                 _logger.LogError("Failed to fetch playlist via GraphQL: {StatusCode}", response.StatusCode);
@@ -801,7 +812,19 @@ public class SpotifyApiClient : IDisposable
                 };
                 request.Headers.Authorization = new AuthenticationHeaderValue("Bearer", token);
                 
-                var response = await _httpClient.SendAsync(request, cancellationToken);
+                var response = await _webApiClient.SendAsync(request, cancellationToken);
+                
+                // Handle 429 rate limiting with exponential backoff
+                if (response.StatusCode == System.Net.HttpStatusCode.TooManyRequests)
+                {
+                    var retryAfter = response.Headers.RetryAfter?.Delta ?? TimeSpan.FromSeconds(5);
+                    _logger.LogWarning("Spotify rate limit hit (429) when fetching library playlists. Waiting {Seconds}s before retry...", retryAfter.TotalSeconds);
+                    await Task.Delay(retryAfter, cancellationToken);
+                    
+                    // Retry the request
+                    response = await _httpClient.SendAsync(request, cancellationToken);
+                }
+                
                 if (!response.IsSuccessStatusCode)
                 {
                     _logger.LogWarning("GraphQL user playlists request failed: {StatusCode}", response.StatusCode);
@@ -851,11 +874,11 @@ public class SpotifyApiClient : IDisposable
                 if (itemCount < limit) break;
                 offset += limit;
                 
-                // GraphQL is less rate-limited, but still add a small delay
-                if (_settings.RateLimitDelayMs > 0)
-                {
-                    await Task.Delay(_settings.RateLimitDelayMs, cancellationToken);
-                }
+                // Add delay between pages to avoid rate limiting
+                // Library fetching can be aggressive, so use a longer delay
+                var delayMs = Math.Max(_settings.RateLimitDelayMs, 500); // Minimum 500ms between pages
+                _logger.LogDebug("Waiting {DelayMs}ms before fetching next page of library playlists...", delayMs);
+                await Task.Delay(delayMs, cancellationToken);
             }
             
             _logger.LogInformation("Found {Count} playlists matching '{SearchName}' via GraphQL", playlists.Count, searchName);

allstarr/Services/SquidWTF/SquidWTFStartupValidator.cs

@@ -73,7 +73,11 @@ public class SquidWTFStartupValidator : BaseStartupValidator
                 {
                     try
                     {
-                        var response = await _httpClient.GetAsync(endpoint, ct);
+                        // 5 second timeout per ping - mark slow endpoints as failed
+                        using var timeoutCts = CancellationTokenSource.CreateLinkedTokenSource(ct);
+                        timeoutCts.CancelAfter(TimeSpan.FromSeconds(5));
+                        
+                        var response = await _httpClient.GetAsync(endpoint, timeoutCts.Token);
                         return response.IsSuccessStatusCode;
                     }
                     catch

docker-compose.yml

@@ -17,8 +17,11 @@ services:
     networks:
       - allstarr-network
 
+  # Spotify Lyrics API sidecar service
+  # Note: This image only supports AMD64. On ARM64 systems, Docker will use emulation.
   spotify-lyrics:
     image: akashrchandran/spotify-lyrics-api:latest
+    platform: linux/amd64
     container_name: allstarr-spotify-lyrics
     restart: unless-stopped
     ports: