This message was posted by a user wishing to remain anonymous
Going back this is some kinda weird ARP bug . I know, I know this is not a typical setup , but lets be honest 20% of all setups are odd balls .
After sitting down and getting some time to put into this .
Using the filter suggested. show sec flow ses src-prefix 10.10.10.0/27 dest-prefix 10.10.10.0/27
I seen half duplex session which I never noticed before..
Session ID: 22603, Policy name: WHAT-IS-GOING-ON/25, Timeout: 14, Session State: Valid
In: 10.10.10.6/47752 --> 10.10.10.10/8007;tcp, Conn Tag: 0x0, If: irb.0, Pkts: 3, Bytes: 180,
Out: 10.10.10.10/8007 --> 10.10.10.6/47752;tcp, Conn Tag: 0x0, If: irb.0, Pkts: 0, Bytes: 0,
The 10.10.10.10 PBS is a physical box on an access port, 10.10.10.6 vm is running on Proxmox on a truck port
Set a static arp on access port 10.10.10.6/27 to vm running on trunk port 10.10.10.10/27
ICMP and Proxmox traffic started flowing! Head slap moment as I dont recall if I set static arps on both sides last when I tested
and what I was testing with . (being in a rush gets you ever time)
Testing using higher secondary networking
Put another device on the Proxmox server 10.10.10.36 and a physical device on 10.10.10.41
Watching wireshark capture
host 10.10.10.36
? (10.10.10.41) at <incomplete> on ens18
Set static arp and seen wireshark show icmp from .36 to .41 but no replies.
host 10.10.10.41
? (10.10.10.36) at <incomplete> on enp1s0
set static arp for .36 and traffic is flowing!
Going to start over and do a format install running 22.4RxsX ..
FYI I did for a brief period set set the subnet mask from /27 on both addresses to /26 which made it 10.10.10.0/26 (0-63),
and still I could not do 10.10.10.6 10.10.10.10 communication .
I do not have any support on the unit and I will be doing my best to keep the community in sync with my progression.
FYI . I was making notes about the trunk / access ports because im using to tshooting DAI / IP SRC Guard where the trunk port is always trusted.
SRX3xx does not support those features.
Guessing .6 had the cache entry as its been up 21days well before the swap from the 320 to 340, and .10 for 3 days ,
Thanks Steve / Gavin .. Have a great Thanksgiving or if your from outside US have a great Friends giving.
Original Message:
Sent: 11-17-2023 14:30
From: Anonymous
Subject: No East West traffic after migrating from SRX320 to SRX340
This message was posted by a user wishing to remain anonymous
Head scratcher.
I copied the config over from 320 to 340 and expected everything to just work.. :( IPv4 to devices on same subnet no longer can talk to each other.
IPv6 still works as I can inter communicate with things on same vlan no problem. :) but it was never setup to have secondary networks .
vlan 10 irb.0 with 2 ipv4 networks
10.10.10.1/27 primary / preferred
10.10.10.49/27 (its not start or end of the subnet but it does not have to be)
Host IP 10.10.10.3 can hit anything in the 10.10.10.32/27 network no issues and vice versa. Hosts can access Internet just fine via IPv4 and IPv6.
However 10.10.10.0/27 to anything other that SRX gw nothing, and hosts in 10.10.10.32/27 same :/
I never get an arp back from hosts and when I static set arp with correct info still nothing.. Its like all traffic gets silently dropped by SRX .
Makes me think I am hitting a bug .. was srx320 was running 21.4R3-S4 . 340 running 22.4R2-S2.6 Had tried to 21.4R3-S5 but had funky other issues.
Just asking if anyone can suggest best version of Junos they run on theirs .. Jtac recommended is 21.4R3-S4
Having a lot of odd issues with jdhcp too, but working on moving that over to a Kea server.
If you read this .. Thanks for your time.