The problem was in Traffic Shaper ..

When I prioritized the traffic on the network using the wizard, I reported the speed of the WAN link, in this case, 20 mbps.

So, as I checked, pfsense limited the network bandwidth to 20 mb (2.5 MB), including LAN and WAN. As this information is not made explicit in pfsense, I only notice when I deleted Traffic Shaper and set it up again.

The solution I found was to report the maximum capacity of the circuit (1 Gb) to use the entire internal network bandwidth. Otherwise, the entire network is limited to the speed of the WAN link.

I had done this configuration for a long time, but since the network was not segmented, the internal traffic was not through pfsense, so when I created vlans, pfsense limited the band ..

Sorry, just to clarify more ...

In fact, I made mistakes in the nomenclatures. The rates quoted above are in MB (not mb), but I have already edited the post.

My network is standard gigabit and I am sharing a single physical NIC among 4 VLANs.

I'm testing my network speed by "iperf"

In inter-host transfers from the same vlan, the rate goes from 100 mbps

Well that sure is not GIG... Gig you should see around 113MBps file transfer for example.

mb = megabits
MB = MegaBytes

So please clarify what your seeing, and exactly what your using to test these speeds.

I take it your uplink on these vlans to pfsense is on the same physical interface. So you have to understand that is a hairpin, and bandwidth on the physical interface is shared between all traffic flowing on that interface... If you have some client downloading from the internet on vlan X, and then you want vlan Y and Z to talk to each other - that bandwidth that vlan X is using eats up the available bandwidth, etc..

So please draw out your actual connections, what exact switch your using and lets see a iperf test between devices on the same vlan and then routed across pfsense. While you take a look at the traffic graphs on pfsense to show that the interface is not being saturated.