Right now I'm running the scaling tests on Xtract (for all of MDF) and it seems like we fall pretty well below 3000 (~1200) requests per 5 minutes by manually changing the number of threads. This is not conducive to unique network conditions.
I propose that we make this internally configurable to m and n% where m% of requests are for sending and n% are for polling.
Right now I'm running the scaling tests on Xtract (for all of MDF) and it seems like we fall pretty well below 3000 (~1200) requests per 5 minutes by manually changing the number of threads. This is not conducive to unique network conditions.
I propose that we make this internally configurable to m and n% where m% of requests are for sending and n% are for polling.