VLANs stop working after while

[Chura]

OPNsense running over Promox, with MellanoxConnectX3 as passthrough
My lan interface is using the native vlan and works great, however it also carry 2 vlans

vlan098: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1500
   description: WifiGuests (opt3)
   options=80000<LINKSTATE>
   ether 7c:55:30:90:ce:e0
   inet 192.168.195.99 netmask 0xffffff00 broadcast 192.168.195.255
   groups: vlan
   vlan: 98 vlanproto: 802.1q vlanpcp: 0 parent interface: mlxen1
   media: Ethernet autoselect (1000baseT <full-duplex,rxpause,txpause>)
   status: active
   nd6 options=29<PERFORMNUD,IFDISABLED,AUTO_LINKLOCAL>
vlan099: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1500
   description: IoT (opt4)
   options=80000<LINKSTATE>
   ether 7c:55:30:90:ce:e0
   inet 192.168.199.99 netmask 0xffffff00 broadcast 192.168.199.255
   groups: vlan
   vlan: 99 vlanproto: 802.1q vlanpcp: 0 parent interface: mlxen1
   media: Ethernet autoselect (1000baseT <full-duplex,rxpause,txpause>)
   status: active
   nd6 options=29<PERFORMNUD,IFDISABLED,AUTO_LINKLOCAL>

those vlans stops working after few min, if i'll restart again they will work for few min and then again stop
I tried to get DHCP when its not working, I've seen that I'm being offerer DHCP from the native vlan
<190>1 2023-06-19T13:48:41+03:00 opn.i.reshef.org dhcpd 52290 - [meta sequenceId="817"] DHCPDISCOVER from cc:66:0a:34:da:78 (ChuraDev) via mlxen1
<190>1 2023-06-19T13:48:41+03:00 opn.i.reshef.org dhcpd 52290 - [meta sequenceId="818"] DHCPOFFER on 192.168.192.147 to cc:66:0a:34:da:78 (ChuraDev) via mlxen1

can it be that tagging is gone after few min ? bug ?

[Chura]

Could it be related ? https://github.com/opnsense/core/pull/4918
not sure if its the same issue as I see

[CJ]

What do you mean that your lan is using the native vlan?  Can you provide some more information and possibly a network diagram?  It's unclear how you have things set up.