Dead Reckoning Summary
The Grand Fleet’s commission office had been open for weeks without registering a single vessel. Captains filed manifests, waited, and heard nothing. Ships commissioned on paper never reached open water. And when cargo did arrive at port, it was wrong. No tool along the route admitted to knowing why.
Dead Reckoning asked players to trace that failure across three layers of a real platform engineering stack: repair the broken Backstage template swallowing every commission before it reached the archives, untangle the misconfigured handoffs between Gitea, Argo Events, Argo Workflows, and Argo CD that left every vessel stalled in the yards, and finally restore the distributed trace that turned a chain of per-tool successes into a readable record of what actually moved from bow to stern, and where the cargo went astray.
Full Leaderboard: July 2026 
| Rank | Username | Solved | Beginner | Intermediate | Expert | Total |
|---|---|---|---|---|---|---|
| 1 | @iw4p | 3 | 95 | 100 | 95 | 290 |
| 1 | @theharithsa | 3 | 100 | 90 | 100 | 290 |
| 3 | @alvin | 3 | 90 | 95 | 90 | 275 |
Your Credly badges and the 50% LF certification voucher are coming your way soon.
The Journey Through Dead Reckoning 
Every commission in the Grand Fleet begins at the office: a captain files a manifest, and the scaffolder is supposed to handle the rest. But the office had been open for weeks without recording a single vessel. The Backstage software template was the first thing to repair. Somewhere in the sequence of fetch, publish, and catalog registration, a step had been misconfigured. Manifests went in and nothing came out. Players traced the parameters, found the broken action, and got the first vessels through the gate.
With the office running, the fleet’s next problem came into view. Commissioned ships were still not sailing. Between the office and open water lay a chain of yards, each automated to pick up where the last left off: a Git push triggering an Argo Events Sensor, a Sensor launching a parameterized Argo Workflow, steps in that workflow calling on the cluster permissions they needed to run, and an Argo CD ApplicationSet waiting at the far end to discover the new repository and sync the vessel into formation. Every yard had something misconfigured. Players moved through them one by one, repairing RBAC settings, correcting event sensor bindings, and getting the full handoff sequence running clean.
The Expert level turned the problem inside out. The pipeline was running. The logs said everything was fine. Captains kept signing for the wrong cargo. The navigation log, the distributed trace meant to record each commission as one unbroken voyage from manifest to harbor, had gone dark the moment a vessel left the office. Without that trace there was no way to see where the data diverged. Players repaired the trace context propagation across the asynchronous boundary, carried the W3C traceparent through the push-triggered pipeline, and used Jaeger to read the complete voyage end to end until the fault surfaced: one step quietly substituting the wrong value, invisible to every per-tool log but unmistakable in a trace that followed the data from start to finish.
By the Numbers 
- 9 solutions submitted across all levels
- 3 engineers solved every active difficulty level
- 89 views across all challenge topics
- 67 total forks since the challenge series launched
What’s Next? 
We are taking a break from new challenges in August. Most of us, as a community, will be away enjoying the summer break.
The site does not go quiet. We are publishing one post a week on tools from the challenges:
- 4 August: Backstage
- 11 August: Kyverno
- 18 August: OpenTelemetry
- 25 August: Argo CD
Dead Reckoning has no published solutions yet. If you solved a level, ADVENTURES.md explains how to add your walkthrough. If you would rather keep it informal, post in the discussion thread for that level instead.
See you in September. 
Closing Thoughts 
Dead Reckoning covered three connected layers of a platform engineering stack: Backstage software templates and scaffolder actions at the front end, a GitOps delivery pipeline wiring Gitea, Argo Events, Argo Workflows, and Argo CD through the middle, and OpenTelemetry distributed tracing at the end, with Jaeger as the lens for reading what actually happened across the full chain. Each level built directly on the previous one. The fleet could not sail until the office was running, and the fault in the cargo could not be found until the trace was repaired.
Three participants completed Dead Reckoning in July, and all three solved every level. Nine solutions were submitted across beginner, intermediate, and expert, covering template debugging, pipeline repair, RBAC configuration, ApplicationSet discovery, async trace context propagation, and fault localization in Jaeger. That is a demanding set of skills to work through in a single month.
Thank you to everyone who sailed with the Grand Fleet this July. The broken logs and mislabeled cargo were fictional, but the instrumentation, the pipeline configuration, and the debugging methods are the same ones that matter in production.