CSE Roadmap

Core tier · about 3 hours · 12 parts

Part 9 of 12 · 2 min read

Mistakes that lose runs or repeat charges

Common mistakes, and what the example does not do.

Mistakes that lose runs or repeat charges

  • Restarting from scratch and repeating spend and effects.
  • Journalling "step done" without the result, so replay cannot continue.
  • No idempotency keys on effects, trusting the journal alone.
  • Non-deterministic orchestration code (time, random, unordered iteration).
  • Resuming old journals against new code.
  • Holding a worker while waiting for a human.
  • Compensations that are not idempotent or never tested.
  • Unbounded journal growth on very long runs.

What the example does not do

About this example What the example does not do
  • In-memory journal, single process: no real storage, ordering, concurrent writers or exactly-once journal writes.
  • Crash positions are injected, not real process kills; partial writes and corrupted journals are not simulated.
  • Probabilities and costs are assumptions.
  • Idempotency is assumed to be honoured downstream; many real APIs do not support keys.
  • No history versioning, patching or migration, only a refuse-to-resume hash.
  • Compensation is a toy: real compensations can fail or be impossible.
  • Optional real experiment: run a real agent under a durable engine or your own journal, kill the worker at random during runs in staging, and compare cost, duplicate effects and completion with a restart-on-failure setup.

Check yourself

A 7-step agent run crashes at step 5. What does restarting from scratch cost compared with resuming from a journal?

Show answer

Restarting cost 12.43 per run against 8.22 for resuming, and it repeated side effects 243 times. Resuming skips completed steps; an idempotency key stops the step in flight from running twice.

What you will use daily at work

Practice What it looks like
Journal each expensive or non-repeatable step Replay reuses recorded results.
Key every external effect by run id and step Retried steps then do nothing twice.
Record time and random values Replay must follow the original path.
Version workflows with runs in flight Keep old versions until old runs finish.