This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

IPSec Site to Site VPN

Hello everyone,

     I have a Sophos XG firewall and SonicWALL SOHO site to site VPN setup between those two firewalls. It connects up just fine and stays connected for hours but randomly it will go down and I have to go manually connect it back up. Luckily it's not a super important VPN but still I don't understand why it randomly goes down. 

I do have the latest firmware installed on both the SonicWALL and the Sophos XG. 

I have changed the dead peer detection on and off both on the SonicWALL and the Sophos XG to see if that makes a difference but it does not matter if that is on or off. Does it matter on the type of encryption I am using between the two? Is there a known issue with this that I am missing?

Any help would be appreciated. 

Thank you. 



This thread was automatically locked due to age.
  • Its really a nightmare!!

    We are running about 800 VPNs tunnels between Sophos XG and several Sonicwall gateways all around the world. We have several support session with the Sophos Support and Global Escalation Team --> currently there is no solution for our issues.

    We are running 16.0.8 MR8 since last week.

    With 16.0.8 MR8 we have IPSec issues with multiple subnets on both sites - we tried everything to get a stable configuration with IKEv1 and different IPSec policys 3DES/SHA-1/None, AES256/SHA2-256/DH2 or DH5, same SA lfetime or smaller in phase2.

    The Support have advise us to install the new 17.0.5 MR5 - we started the update... (I didn´t do this again without a test-environment!)

    Issues:
    - IKEv1 with multiple subnets on both sites --> workaround switch about 800 tunnels to IKEv2 (would be a long night!)
    ... after we have everything migrated, we thought that we have it stable now... next morning 50% of tunnels were down!

    - WebGUI will hang and the appliance will get unresponsable --> failover to the slave appliance.
    ... All IPSec connections must be actived manually cause the failover doesn´t switch the status (enabled/disabled)...

    Support have tried to find out the root cause for our problems and tried some fixes.

    - WebGUI becomes serveral times unresponsable... we must had restarted the IPSec connections over 7 days.
    ...

    Today we decided to change the plattform to SG. Today we have configured everything from scatch new - we are stable but running an old plattform now... very poor!!!

     .-(

  • Sven, 

    Check out the link I shared in a previous post on this thread. 

    The other week I spent a week working with Sophos and the escalation team to sort all these issues. I was seeing all the same stuff. 

    Long story short, in V17 they rebuilt the entire VPN daemon. This meant that settings on V16 which had worked, sent my firewalls into a constant loop trying to establish vpn connections. I would recommend turning off non vital tunnels for an upgrade to 17.x. Doing that will fix the GUI hanging issue. 

    I would be willing to bet most of your issues are related to DPD and keying times. There are detailed settings in the link I shared, but basically make sure one side of the tunnel has DPD set to disconnect, ensure Phase 1 is longer than phase 2.  Turning off PFS (not ideal) can solve some rekeying issues. 

    I do have XG v17 tunnels with stable connections to Sonic Walls but it took some work.