Closed jaredhirsch closed 11 years ago
ideally we can distinguish between individual types of 4xx errors, eg CSRF failure/malformed request 400s vs relatively benign 404s
can't be done--not better finer-grained monitoring of HTTP codes, anyway. we can query for instance health, but doesn't seem worth it. reopen if it is.
Talked with @vmunix about this last night, he suggested taking a look at how campaign monitor watches ELBs. Maybe we can get more granularity than 1 minute.