HTTP reliability laboratory / 01

See what your retry code actually does.

Generate a controlled HTTP endpoint. Send real requests. Observe timing, backoff, Retry-After handling, and idempotency without digging through application logs.

Create a test lab

Anonymous · temporary · automatically deleted

Observed attempt pathExample
POST/test/••••••••
  1. 01500+0 ms
  2. 02500+1,024 ms
  3. 03200+2,047 ms

What RetryLab reveals / 02

A purpose-built view of the space between attempts.

Retry behavior lives across request logs, response timestamps, and client settings. RetryLab turns those scattered signals into one controlled experiment.

01
Response completion to next request
02
Configured Retry-After versus observed delay
03
Hashed idempotency-key consistency

Configure an experiment / 03

Choose a known failure pattern or define your own sequence.

Every lab receives separate endpoint and dashboard capability tokens. No account is required.

API contract v1

Loading scenarios from the RetryLab API…

After final step

Creating a lab generates two separate bearer tokens. Do not include production secrets in test requests.

Operating sequence / 04

From configured failure to observable evidence.

  1. 01 / Configure

    Define the response path.

    Select a preset or arrange up to ten status, delay, header, and body steps.

  2. 02 / Connect

    Point your client at the lab.

    Copy the generated endpoint into the code whose retry policy you want to test.

  3. 03 / Observe

    Compare intent with behavior.

    RetryLab records sanitized attempts and calculates only what the server observed.

Temporary infrastructure / 05

Built to test disposable traffic, not hold production data.

Labs expire after 24 hours. Sensitive request headers are never stored, bodies are sanitized and size-limited, and raw capability tokens are not persisted by the API.

Read the data-handling notes