Another likely OpenVPN(UDP)+SOCKS5 bug, likely unrelated to previous (connection "stalls/stutters" occasionally, with very inconclusive logs)

Describe the bug
A clear and concise description of what the bug is.

So, suppose you have OpenVPN(UDP) and a localhost SOCKS5 proxy (I've only have witnessed this behavior on localhost SOCKS5 but because it's hard to reproduce it might be that simply I was not patient enough with remote one)

OpenVPN(UDP)=SOCKS5(localhost)=[Remote another machine]=[Remote OpenVPN Server]
OR
OpenVPN(UDP)=SOCKS5(localhost)=[another machine on same LAN]=[OpenVPN Server on same LAN]

Connection can run for a long time, and then suddenly staggers, but not long enough for it to timeout (think 4-15 seconds)

Nothing happens in the logs of both SOCKS and OpenVPN client at this point

Then suddenly connection resumes but the log of OpenVPN spews a bunch of errors about possible replay.

Happens when SOCKS5 is a shadowsocks sslocal + some sserver AND happens when SOCKS5 is a vanilla SOCKS5 (HEV used for this test)

To Reproduce
See above

Expected behavior
For connection to run without hiccups

Version information (please complete the following information):

• OS: Debian 11 (VM)

• OpenVPN version:
  OpenVPN 2.6.5 x86_64-pc-linux-gnu [SSL (OpenSSL)] [LZO] [LZ4] [EPOLL] [PKCS11] [MH/PKTINFO] [AEAD] [DCO]
  library versions: OpenSSL 3.1.1 30 May 2023, LZO 2.10

Due to the fact that the bug takes a long while to reproduce and doesn't cause conclusive log evidence I've only tested it on one configuration (but with two different completely unrelated SOCKS5 server options)

Additional context
Add any other context about the problem here.

Start by reproducing the OpenVPN(UDP)=SOCKS5 topologies described on Debian 11, using both the shadowsocks sslocal setup and HEV SOCKS5 server. Compare OpenVPN and SOCKS5 behavior during the 4–15 second stall, especially the subsequent possible-replay errors. Done means identifying a reproducible cause and a verified fix or a conclusive diagnosis.