0

2 of my PCs are connected to my router like so:

Both PCs' ethernet cables -> Switch -> ethernet cable -> ethernet socket -> ethernet cable from socket -> switch -> ethernet cable to socket -> ethernet socket -> ethernet cable -> Router (Asus RT-AX92U)

All the ethernet cables are at least Cat-5e (1 gbps). All switches are unmanaged.

Every once in a while (maybe once every 2 hours or so?) this is what I get when I ping my router:

Reply from 192.168.1.1: bytes=32 time<1ms TTL=64
Reply from 192.168.1.1: bytes=32 time<1ms TTL=64
Reply from 192.168.1.1: bytes=32 time=1ms TTL=64
Request timed out.
Request timed out.
Request timed out.
Request timed out.
Reply from 192.168.1.1: bytes=32 time<1ms TTL=64
Reply from 192.168.1.1: bytes=32 time<1ms TTL=64
Reply from 192.168.1.1: bytes=32 time=1ms TTL=64

When this happens, I notice that I can still hear everyone perfectly fine in the Discord voice call throughout the timed out period, but they cannot hear me at all. I also get updates to other characters in the online game I play, just that the game will not register my own character's commands until the timed out period ends.

BOTH PCs always fail in this same manner at the same time.

This somehow makes me feel like incoming UDP packets keep coming through fine and it's just TCP packets or upstream UDP packets that do not? Which means the issue is anything upstream doesn't work (since TCP requires upstream ACKs I believe).

But this baffles me as to why this happens, I have never experienced anything like this where ethernet failed me in this weird way where downstream UDP packets still work perfectly fine. And I have no idea how to easily debug this (eg. connect straight to router, connect just to first switch) since I cannot easily replicate this issue and I cannot physically access some of these wires as they are inside my walls.

What is causing this and what do people recommend I do to fix this? Is there a specialist ethernet testing device I can use that will easily and consistently tell me exactly where this is failing? Has anyone encountered something similar before?

Improve this question
10
  • Just for a start, are there any 100Mbps Ethernet links involved or are all links 1Gbps? (The two have significant differences in the way bidirectional transmission is done – 10/100Mbps Ethernet has separate up/down pairs, 1Gbps doesn't – so e.g. a failure of one wire pair would have different results depending on the link speed.)
    – grawity_u1686
    Commented Jan 28 at 15:13
  • Are any of the switches 'managed' (i.e. do they have their own IP addresses) or are they all unmanaged? Being able to ping a switch would make it easier to narrow the problem down. (Often managed switches even have a rudimentary "cable test" function...)
    – grawity_u1686
    Commented Jan 28 at 15:15
  • @u1686_grawity thanks for that, addressed both comments as an edit on the original question.
    – Molten Ice
    Commented Jan 28 at 16:55
  • You say the cables are Cat5e, but that in itself doesn't guarantee that 1Gbps will always be negotiated between devices (whether due to sub-spec cable, or due to damaged wire pairs, or damaged connectors, or manual speed settings...) It would be useful to check the actual links as reported by the PCs/routers/switches (Windows shows the link speed in Get-NetAdapter, switches often have a distinct LED color, etc).
    – grawity_u1686
    Commented Jan 28 at 17:32
  • Ah I see, thanks for that I can confirm running Get-NetAdapter gives link speed 1 Gbps. That should be all the way to the router right?
    – Molten Ice
    Commented Jan 28 at 18:49

0

Reset to default

You must log in to answer this question.

Browse other questions tagged .