investigate-stuck-messages

Diagnose stuck relayer queue messages by correlating retry data and error reasons.

72|602|Updated Dec 13, 2021
One-click install
npx skills add https://github.com/hyperlane-xyz/hyperlane-monorepo --skill investigate-stuck-messages
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: investigate-stuck-messages
Source: https://github.com/hyperlane-xyz/hyperlane-monorepo/tree/main/.claude/skills/investigate-stuck-messages
Command: npx skills add https://github.com/hyperlane-xyz/hyperlane-monorepo --skill investigate-stuck-messages

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Investigate stuck messages in the relayer queue by querying the relayer API, correlating retry data, and surfacing message IDs for denylisting when alerts indicate queue length or stuck prepare messages.

Core Features & Use Cases

  • Grafana alert URL workflow: automatically extract app_context and remote labels and drive the full investigation.
  • Manual workflow: provide app_context and remote to run targeted investigations.
  • End-to-end data correlation: fetch domain IDs, query the relayer API, filter messages by recipient, and fetch actual errors from GCP logs.

Quick Start

Use this skill with a Grafana alert URL to auto-detect app_context and remote, or provide app_context and remote directly to start an investigation. Review the results to identify full message IDs for potential denylisting and root causes.

Frequently Asked Questions about investigate-stuck-messages

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
Why are my relayer queue messages stuck and how do I investigate the errors?

To investigate stuck relayer queue messages, query the relayer API to correlate retry data and fetch actual errors from GCP logs, which surfaces full message IDs and actionable steps for denylisting.

How do I investigate stuck relayer messages directly from a Grafana alert URL?

Provide a Grafana alert URL to automatically extract app_context and remote labels, driving the full stuck message investigation workflow across the relayer API, domain metadata, and GCP logs.

Can I manually query relayer queue issues without a Grafana alert URL?

Yes, you can manually provide app_context and remote inputs directly to run targeted relayer queue investigations that fetch domain IDs, query the relayer API, and extract GCP log errors without an alert URL.

What is the process to extract message IDs for denylisting from the relayer API?

Extracting message IDs for denylisting involves querying the relayer API to fetch domain IDs, filtering messages by recipient, and correlating retry data with GCP log errors to surface full message identifiers.

Does relayer queue investigation work on mainnet3 deployments?

Yes, relayer queue investigation supports both Grafana alert URL and manual workflows across mainnet3, consolidating relayer API data, domain metadata, and GCP logs to present retry counts and full message IDs.

How do I find the root cause of stuck prepare messages in the relayer queue?

Find the root cause of stuck prepare messages by correlating relayer API retry data with actual errors fetched from GCP logs, which surfaces specific error sources and full message IDs for denylisting.