Join the Conversation

To sign in, use your existing MySonicWall account. To create a free MySonicWall account click "Register". - Appliance crashes when Exchange Application Offload enabled

BWCBWC Cybersecurity Overlord ✭✭✭


a couple days ago I updated a SMA 500v from to because the customer would like to benefit from the newly implemented WireGuard protocol. It's a small deployment, 5-10 Users, offloaded Exchange 2016 and remote Access via NetExtender. Running for years without major problems.

But immediately after the update, the Appliance reboots every few minutes without obvious reason. After some digging I was able to narrow the problem down to the Exchange Portal. If I make the Exchange Portal inaccessible, the Appliance stays on, but with enabled Exchange Portal after a few minutes the system crashes. The console shows what is going on and the watchdog forces the system to reboot. I tried to disable all users and put them back on one by one, but it's not related to a specific connection.

Did anyone experienced something similar? I already recreated the Portal and tinkered with the settings, but nothing helped.

I downgraded to and since than the appliance is back to normal.

@Simon did you heard about that before?


Category: Secure Mobile Access Appliances

Best Answer

    BWCBWC Cybersecurity Overlord ✭✭✭
    Answer ✓

    A quick update on this, the newly released Firmware seems to be working much better and the Appliance is no longer crashing (15 hours in and counting).



  • VaiVai SonicWall Employee

    If the initial configuration was upgraded from 8.x firmware, it can have certain characters/configurations which will result in such failures. I would also remove the certificates for the portal and check.

  • BWCBWC Cybersecurity Overlord ✭✭✭

    @Vai I checked in status.txt and the earliest firmware seems to be 9.0.0, which was IMHO a necessary platform upgrade because older version were not able to upgrade via .sig file. The Customer Deployment by itself dates back up to 2012 which brings us in the SRA 6.x ballpark or even earlier.

    I cannot tell for sure if I ever created the config from scratch on that path.

    But the main question is, why is working fine and freaks out. This shouldn't be caused by some characters or even the Cert which is working fine for VirtualOffice?

    Because of the extensive impact I see no way to handle this with Support, because the Customer would be highly impacted by that.


Sign In or Register to comment.