I have two trace files captured at different ends of a transaction with what is obviously a firewall in between that scrambles the sequence numbers. As a result when I perform a merge, it sees the same two packets from either end as different and stitches instead of merging them. As the sequence number changes, you could not use the TCP CRC either to identify the packets, however if you have a clean trace from either end that is just monitoring one connection you could perform a merge based on the same socket.
Is there any of doing this in TT. I loathe the thought of go through each packet pair and manually add the latency times together.
Solved! Go to Solution.