David Bar-On

> The `MSG_TRUNC` seems to affect TCP and UDP different. You have things set-up in a way that I would expect only TCP throughput to be improved. Has anyone seen...

> If I understand the Linux kernel documentation correctly, recv(fd, buf, nleft, MSG_TRUNC) will discard different parts of the kernel's buffer depending on if it is UDP or TCP. ......

The changes to support SKIP-RX-COPY where moved to PR #1717, and this PR will be used only for the support of zero-copy using SO_ZEROCOPY/MSG_ZEROCOPY. > The math for Nread/Nrecv took...

Added a separate PR #1720 for the support of MSG_ZEROCPY/SOZEROCOPY, now with notifications support. Closing this PR as its functionality is now split between PRs #1717 and #1720.

@Madhu-api, what was the the IP address used by the client before the rout change? Was it 172.21.0.3 as after the change or a different IP address was used? Also,...

@Madhu-api, can you send the client's IP address in the two routing options? In the original issue text you show that before the route change the client's IP address was...

Re-implementation for the Multi-thread iperf3. I am not sure whether the original problem still happen with multi-thread. **Some of the changes may be worth implementing in any case**, so if...

@RizziMau, I also suggest to run the **server** using **`--debug=3`** option. The debug messages may help to better understand were the server got stuck, by comparing successful run output to...

> I'm seeing these compilation errors. Am I missing a step? @bprodoehl,, did you re-run `configure` (and maybe `bootstrap.sh`)? This should be done as these are probably new declarations in...

@PreetiMSobarad, can you also list the server's iperf3 command? Also, I suggest to add the following to the client's (and maybe also server's) command: - _--debug_ and _--verbose_ to get...