Hello friends!
I have the following scenario:
- One HPE JL254A 48 port switch being used as core L2 and L3 concentrator: sw01
- One HPE JL682A 24 port switch being used as access device: sw02
I have configured some VLANs with jumbo frames enabled.
There is a fiber optic transceiver connected to sw01 port 11. This fiber leads to a transceiver connected to sw02 port 24.
I have configured VLAN 4000 on both switches. Hypothetic IP addresses are shown below:
- sw01: 10.0.127.1/30
- sw02: 10.0.127.2/30
There is a user computer named dev01 connected to VLAN 3999 on sw01, with ip address 10.0.0.1/24. VLAN 3999 is assigned IP address 10.0.0.254/24 and is the default gateway for dev01. Jumbo is not enabled for VLAN 3999, but is enabled globally on JL254A and some other VLANs.
There is another user computer named dev02 connected to VLAN 4001 on sw02 with ip address 10.0.64.1/24. VLAN 4001 is assigned IP address 10.0.64.254/24 and is the default gateway for dev02. Jumbo is also not enabled for this specific VLAN 4001, but is enabled globally on JL682A switch.
A user on dev01 is trying to open UltraVNC connection to dev02, but VNC connection drops constantly making it impossible to support user on dev02.
I made a Wireshark capture and I have noticed that there countless TCP retransmissions of VNC port 5900.
What can be wrong? Why there are so many retransmissions of VNC packets sent between dev01 and dev02?