Analyze a capture¶
Most investigations start here. These commands give you the shape of a capture, rank what looks wrong, and let you narrow in. Once you know which kind of problem you have, move to Repeats, Failures or Inspect.
Start with one of three¶
| Command | Use it when | Page |
|---|---|---|
summary |
You want the shape of the capture and a short list of next steps. It runs when you give no command. | Summary |
auto |
You want WireTrail to run the follow-up analysis for the serious findings too. | Auto |
diagnose |
You want only the ranked findings, with their evidence. | Diagnose |
All three rank findings with the same logic, so they never disagree. The difference is how much they print.
Look around¶
| Command | Answers | Page |
|---|---|---|
hosts |
Which servers, how many requests, how fast, how many errors | Hosts, subsystems and endpoints |
subsystems |
The same, grouped by integration or owning team | Hosts, subsystems and endpoints |
endpoints |
Which routes were called and how they answered | Hosts, subsystems and endpoints |
timeline |
What happened, in order | Timeline |
validate |
Whether the capture holds what the analysis needs | Validate a capture |
Narrow it¶
Every command takes --filter to look at a subset, and --top to bound the length of its lists.
validate, show-entry and curl <id> ignore --filter. Filter entries covers the clauses.
Recipes¶
- Triage an unfamiliar capture is the order of commands to run when someone hands you a capture and a vague complaint.
Related¶
- Choose the right command starts from your question.
- How findings are ranked explains the severity levels.