1 | initial version |
At which point did you take the trace at server side I guess, too. The session at all looks strange. I would guess there is something inside the oder packet which causes the application to crash.
Another hint is that before session finally is initiated the SYN gots often an RST as an answer.
2 | No.2 Revision |
At which point did you take the The trace at server side I guess, too.
The session at all looks strange.
I would guess there is something inside the oder packet which causes the application to crash.
Another hint is that before session finally is initiated the SYN gots often an RST as an answer.
3 | No.3 Revision |
The trace at server side I guess, too.
The session at all looks strange.
a little bit strange in some details.
But I would guess there is something inside the oder packet which causes the application to crash.
Another hint is that before session finally is initiated the SYN gots often an RST as an answer.
4 | No.4 Revision |
Answer for Trace Update1:
The trace at server side I guess, too. The session at all looks a little bit strange in some details.
But I would guess there is something inside the oder packet which causes the application to crash.
Another hint is that before session finally is initiated the SYN gots often an RST as an answer.
5 | No.5 Revision |
Answer for Trace Update1:trace of Update1
:
The trace at server side I guess, too. The session at all looks a little bit strange in some details.
But I would guess there is something inside the oder packet which causes the application to crash.
Another hint is that before session finally is initiated the SYN gots often an RST as an answer.
6 | No.6 Revision |
Answer for trace of Update1
:
The trace at server side I guess, too. The session at all looks a little bit strange in some details.
But I would guess there is something inside the oder packet which causes the application to crash.
Another hint is that before session finally is initiated the SYN gots often an RST as an answer.
Answer for Update2
traces:
First of all we see differences in the 3way-Handshake of client side and server side.
Handshake at client
side:
- Client advertises 1460 MSS
- Server advertises 1460 MSS
Handshake at server
side:
- Client advertises 1398 MSS
- Server advertises 1460 MSS
Paket 325-330, are to big for the tunnel, and didn´t make it through the tunnel. At the end the client resets the session. Then we must change to the server side trace as the trace is longer. After that resets a few session retries happen and in the end the client tries a session with Fragmentation allowed. Which mostly won´t work well on tunnels.
So my recommendation is: Please try to advertise an adjusted server MSS to to the client. Like the client does on Server side trace.
Some routers are able to do so. see here: https://www.cisco.com/c/en/us/support/docs/ip/generic-routing-encapsulation-gre/25885-pmtud-ipfrag.html
Here you can find an explanation about MSS in general: https://crnetpackets.com/2016/01/27/the-relation-between-maximum-transmission-unit-mtu-and-the-maximum-segment-size-mss/
7 | No.7 Revision |
Answer for trace of Update1
:
The trace at server side I guess, too. The session at all looks a little bit strange in some details.
But I would guess there is something inside the oder packet which causes the application to crash.
Another hint is that before session finally is initiated the SYN gots often an RST as an answer.
=================================================
Answer for Update2
traces:
First of all we see differences in the 3way-Handshake of client side and server side.
Handshake at client
side:
- Client advertises 1460 MSS
- Server advertises 1460 MSS
Handshake at server
side:
- Client advertises 1398 MSS
- Server advertises 1460 MSS
Paket 325-330, are to big for the tunnel, and didn´t make it through the tunnel. At the end the client resets the session. Then we must change to the server side trace as the trace is longer. After that resets a few session retries happen and in the end the client tries a session with Fragmentation allowed. Which mostly won´t work well on tunnels.
So my recommendation is: Please try to advertise an adjusted server MSS to to the client. Like the client does on Server side trace.
Some routers are able to do so. see here: https://www.cisco.com/c/en/us/support/docs/ip/generic-routing-encapsulation-gre/25885-pmtud-ipfrag.html
Here you can find an explanation about MSS in general: https://crnetpackets.com/2016/01/27/the-relation-between-maximum-transmission-unit-mtu-and-the-maximum-segment-size-mss/