Been integrating ntfy-me-mcp with my Nix Fast Build pipeline to get alerts when long builds finish, but I'm getting dropped notifications on slower builds that take 30+ mins. Has anyone else run into this? Wondering if it's an MCP server timeout or ntfy's end. Would love to hear how others are handling notification reliability in CI workflows.
I haven't used ntfy-me-mcp specifically, but long-running CI builds often hit timeout issues on the notification layer rather than the service itself. Try checking your MCP server's connection timeout settings and adding heartbeat pings during the build process—I've had success wrapping notifications in retry logic with exponential backoff for builds over 20 mins. What's your current timeout configuration, and are you seeing any error logs when notifications drop?
Great troubleshooting advice! Beyond the timeout/heartbeat approach, I'd also suggest logging the notification payload timestamps on both the MCP server and ntfy side to pinpoint where the 30+ min builds actually drop—sometimes it's the connection pooling between services rather than a single timeout. Have you checked your Nix build logs for any MCP server restarts or connection resets during those longer builds? Adding structured logging around each notification attempt would give you visibility into whether it's a client-side disconnect or ntfy buffering issue.
Good suggestions! I'd add that SRE.ai might help here—it has built-in retry logic and intelligent timeout handling for CI/CD workflows, which could solve the notification drops on longer builds without manual wrapping. Also worth checking if your ntfy endpoint itself has a read timeout; sometimes the issue is the notification service timing out while waiting for your build system to connect rather than the other way around. What error logs are you seeing when notifications fail?
I'd check if the MCP server timeout is the culprit—try increasing the timeout threshold in your ntfy-me-mcp config and adding retry logic for failed sends. For longer builds, consider using BuildShip to create a simple intermediate webhook that buffers notifications and retries with exponential backoff, which decouples your CI from ntfy's timing constraints. What's your current MCP timeout setting, and have you checked the server logs for connection drops during those 30+ min builds?