Info-109 DHCP Client not running - restarted

Hi, Simon, and welcome to the UTM Community!

You might give #7.7 in Rulz a try.  Any luck with that?

Cheers - Bob

Hi Bob

Still the same problem.

I know this thread is old, but I have had the same problem on my home UTM for some time. I am not sure which update started it. It happens several times a week.

Bump. It is still happening on my home UTM (virtual machine) but not a hardware UTM. I get an email message from the home UTM saying DHCP Client not running - restarted. 

Are you saying that you tried #7.7 and that didn't help?  If using ESXi, change the NIC to VMXNET3.  What if you use a real NIC instead of a virtual NIC? (Sorry, I don't do much VM, so I'm certain I'm not using the correct terminology.)

Cheers - Bob

I started having this issue (Info-109 DHCP Client not running - restarted) soon after updating to 9.600-5.1 UTM (home license) on January 8th.
Sophos runs as 64-bit, on a SuperMicro Atom D525 MiniITX motherboard with 8GB ram and with Intel 82574L NICs on the motherboard for both eth0 and eth1.

Initially after a reboot, the DHCP Client would restart about once or twice a day, but would get more frequent, at about once every hour or two if I left the system running for over a week.
Each DHCP client restart would cause 15-30 second outage where Sophos had no public internet IP address, and thus couldn't route anything to the internet.
This would cause any active TCP connections to drop.  Some UDP based protocols (my VPN to the office) had enough retires to survive the Sophos outage some of the time.

Over the course of several weeks, I ruled out my hardware (I have a backup machine, exact same SuperMicro motherboard) by loading a fresh 9.600-5.1 Sophos image from the ISO media, and swapped the Ethernet cable to my Verizon FiOS Optical Network Terminal (a Nokia ONT model), and also tried a small Gigabit Ethernet switch between router and FiOS ONT (in case signal levels or media sense were a trigger).
I tried auto-negotiation vs 1g full duplex settings for the NIC per Rulz #7.7. I rebooted the FiOS ONT as well many times.
None of this made any difference.

I also upgraded to the recent (Feb 11) Up2Date beta package for 9.601, and still experienced the issue.

I then downgraded Sophos UTM to version 9.510-5.1 on my original machine (fresh install via ISO media), and restored a 9.5 backup configuration from January, and put it into service.

Its been running 5 days on 9.510-5.1 now without a single DHCP client restart !

Btw.. in the System log on 9.510-5.1, I see Sophos UTM doing a DHCPREQUEST every 50-60 minutes, and getting back the same IP from Verizon FiOS with the next renewal between 2700 and 3500 second (45-58 minutes ).  These Sophos initiated refresh timing intervals (50-60 min) might be the trigger points where the DHCP client was crashing on version 9.6 and above, since it would often reset every hour or two.  

I am avoiding updating to any 9.6 or higher version until I see release notes where this issue is fixed.

Hey Mike - welcome to the UTM Community!

Your first post here is a contribution AND it's well written.  I'm looking forward to your continued participation.

Cheers - Bob

Additional info...
 I hooked up an Ethernet tap in between Sophos box and FiOS ONT, so I could watch the raw DHCP data on the wire via WireShark.

Verizon's DHCP Offer have a lease time of 7200 seconds (2 hours)... so Sophos trying to renew DHCP at 50-60 minutes is appropriate for my ISP (i.e. for being about half the lease time)

To put the Ethernet tap in place, I had to briefly disconnect the Ethernet cable, and the DHCP client restarted in Sophos 9.510 .. but this was a purposeful media-sense disconnect/reconnect event.

(This explains the DHCP renew timings, but not the DHCP stop/restart that happens with Sophos 9.600+ at many DHCP renew intervals).