I have client and server applications on 2 different servers in intranet, time taken from the server to the client Could? (I can not rely on machine timestamps because they can be closed)
Split the time reported from two This is a very good, though there is little optimistic estimation of practical latency. (Ping uses the ICMP protocol, you will get different measurements with the TCP application, because TCP is the first one needed.)
If you want to do it in the C ++ application, your best The option can be to call ping command and parse its output Esp. On Unix, where you usually need
root privileges to send an ICMP packet.
Comments
Post a Comment