2 points | by naiks1214 11 hours ago ago
1 comments
Debugging incidents from logs is usually painful, especially when it spans multiple files.
I made a simple CLI tool that tries to figure things out based on things we already do — like looking for ERROR logs, time windows, and some context.
You give logs (even from multiple services), and it:
* builds a timeline * highlights important parts * gives a rough root cause
Example:
`sydes incident analyze api.log worker.log gateway.log --query "Why are session lookups failing?"`
Works best when logs are from multiple services (api + worker etc), where context is spread out.
Still early. Would like to know:
* Does this make sense for you? * Would you trust something like this? * Any improvements you would like to suggest?
Debugging incidents from logs is usually painful, especially when it spans multiple files.
I made a simple CLI tool that tries to figure things out based on things we already do — like looking for ERROR logs, time windows, and some context.
You give logs (even from multiple services), and it:
* builds a timeline * highlights important parts * gives a rough root cause
Example:
`sydes incident analyze api.log worker.log gateway.log --query "Why are session lookups failing?"`
Works best when logs are from multiple services (api + worker etc), where context is spread out.
Still early. Would like to know:
* Does this make sense for you? * Would you trust something like this? * Any improvements you would like to suggest?