QueueForge Client Dead-Letter Queue Onboarding (Onboarding)

QueueForge Client Dead-Letter Queue Onboarding (Onboarding)

1. 01

   Create the master QueueForge account and connect a test RabbitMQ or Kafka cluster

   Validate the full feature set before touching client infrastructure: dead-letter queue visualization, alert creation, message retry, and dynamic rule building. The Free tier gives 100% of platform features with unlimited RabbitMQ clusters for 6 days, which is enough to rehearse the workflow without billing the client.

2. 02

   Map the client's queue architecture and identify the top 3 critical message queues

   Document exchange names, routing keys, and which queues feed the dead-letter exchange. This mapping becomes the scope boundary for the $2,250 SMB Starter Setup, which covers dashboards for exactly 3 queues.

3. 03

   Configure dead-letter queue dashboards for those 3 queues

   Build one dashboard per queue so the client's on-call engineer sees failure volume and routing paths without navigating the full cluster view. Keep the dashboard names aligned to the client's service names, not QueueForge defaults.

4. 04

   Set real-time email and Slack failure alerts with severity thresholds

   Create separate alert rules for stuck consumers, queue growth spikes, and ack stalls. Set thresholds against the client's normal baseline, not generic defaults, so a 200-message spike on a low-volume queue pages someone while a 200-message spike on a high-volume queue does not.

5. 05

   Build dynamic rules for automated recovery

   Use the rule engine to define conditions and actions: reroute messages to an alternative exchange, trigger retries, or fire a webhook into the client's incident tooling. Start with retry-only rules for the first two weeks, then add rerouting once the client trusts the failure classification.

6. 06

   Write the client-facing runbook for interpreting alerts and triggering manual recovery

   Cover what each alert means, which dashboard to open, and when to escalate to the agency versus retry in QueueForge. This runbook is a named deliverable in the SMB Starter Setup and is what justifies the 20-hour setup fee.

7. 07

   Hand off to the client team and schedule the first weekly health check

   Walk the on-call engineer through one live alert and one manual retry in QueueForge. Book the recurring review before the call ends so the monitoring retainer has a calendar anchor.