ECP load testing

[Cantor, Scott]

On 7/9/14, 2:36 PM, "Matthew Slowe" <M.Slowe at kent.ac.uk> wrote:
>
>I think it sees the level of activity that it does for a few reasonsŠ we
>have a *lot* of users using "Smartphones" and routinely using
>ActiveSync/IMAP. Of these, I've observed that ActiveSync seems pretty
>well behaved (start up a session and hold it open for quite a while) but
>IMAP is a lot more "chatty" (given "push" IMAP [IDLE], I'm a bit
>surprised by this) so I suspect a lot of this "traffic" is IMAP clients
>(but I have no real evidence for this).

I always forget people are often talking about Office 365, which is
technically ECP but means really a whole different kind of use case.

You can assume that ECP requests are going to be essentially the same cost
to process as any other login, usually a bit less. So it's not really
about ECP, but your overall traffic load and what that increase will be
from the mail clients. But for load testing purposes, just hammering the
regular endpoints at that expected load should be good enough.

-- Scott