[SOLVED] How to reduce latency of internal bridge connection between two VMs

[piddy]

I have two VMs on the same internal bridge (post about it here); and at the moment when I ping one vm from the other, the latency is in the 100s of ms. I was really surprised by this (the switch is not physical, and communication between the VMs is happening on the same physical computer, so I expected very low latency). Is there a way for me to improve/reduce the latency?

System:
Motherboard: ROMED8-2T
Processor: EPYC 7262
RAM: DDR4, 2400 MT/s RDIMM (64GB)
Boot disk: zpool of 2 x 128GB NVMe m.2 SSDs

The two VMs are:
VM100 - TrueNAS Scale
VM110 - debian server (going to be used to run/manage docker containers

VM100 has one of the ROMED8-2T 10GbE NICs passed through to it. This is set up on a different subnet (192.168.X.X/24). It is also connected to my internal bridge ('media').
VM110 has two bridges vmbr0 (for access to the other subnets on my LAN as well as the WAN via my router); and the 'media' internal bridge.

Any help would be gratefully accepted.

Best wishes,

Philip

 

[piddy]

Apologies - I'd misread the output of ping. The latency is incredibly low... below 1ms. When I read the output, I missed the leading '0.' and thought I was reading latencies of hundreds of ms, not hundreds of microseconds.