From: Ankit M. <ank...@gm...> - 2024-04-03 17:28:46
|
hello, wondering if anyone got a chance to have a look at this. Thanks On Wed, Apr 3, 2024 at 12:24 AM Ankit Mehta <ank...@gm...> wrote: > Hello all, > We are using quickfixj to receive trade execution reports for various > venues. > We are interested in capturing certain network latency stats which is > ideally the difference between the venue sending time which is stamped in > tag 52 and the time message actually hit our application. > > Currently we stamp it under the fromApp method of the quickfixj > application. > > The question I have is whether there is a better alternate way to capture > the message received time somewhere in the socket layer, which can give us > a more accurate picture of network latency. > > When I ran the application in DEBUG mode, I noticed that in the > FixMesageDecoder class when the message is first parsed, the time at which > the parsed message is printed is much earlier than the time it hits the > fromApp method. > > That's why it's important for us to capture the receive time as early as > possible so that we can figure out the network latency much more > appropriately if not accurately. > > Eagerly looking forward to hearing some thoughts from you on this. > > Regards, > Ankit > |