Diagnosing server error 400

[J. Landman Gay]

I need help figuring out why AWS is returning an error 400 "bad request" 
in a limited number of cases. We have a test group of some dozens of 
people and only 2 have the problem.

What I know:

The data sent to the server is the same for everyone, and is correctly 
formatted. I have logs of that and the requests are okay.

One person moved to another computer (and consequently another network) 
and the problem resolved itself.

Once, my client was getting the same 400 error all morning and then that 
afternoon it started working again by itself, without changing anything.

This definitely seems network-related to me but I don't know what it 
could be. The queries are reaching AWS and it is returning the error. We 
don't have access (or don't know how) to see the server logs that Amazon 
has; we only have access to the logs on our own server, and the requests 
never get that far.

Any ideas?

-- 
Jacqueline Landman Gay         |     jacque at hyperactivesw.com
HyperActive Software           |     http://www.hyperactivesw.com