Skip to main content

Fix Windows DHCP Issues In Parallels?

(Here's one weird trick to get it working)

Problem and Solution

For me, this is something of a blast from the past.  I first had Windows DHCP issues using a Windows 8 Tablet, on an unreliable hotel connection in Zim, quite some time ago.  I've never had any real issues since, until I got Parallels up and running, on my M1 MacBook Air, with a Windows Insider ARM build. 

The problem looks like this:


Instead of the typical 192.168.x.x address we get from a normal router, there is is this self assigned 169.254.x.x address and and, from an IPv4 perspective, the computer isn't on the network.  Fortunately, what worked for me on Windows 8, all those years ago, works today on Windows 10.  It is this:

Firstly goto settings and then Network and Internet Settings:


Then continue as indicated below:





Now we enter our IP details, manually.  The settings below are for a BT Smart Hub where the lower range of addresses is reliably ignored by the router's DHCP function.  Finding out what manual settings you can use?  The IP information is available in the Mac's network settings (an iPad is another source of easy to get at networking info).


Finally, we simply repeat steps 1 through 4 but, this time, select Automatic (DHCP) at step 4.  When I do this, Windows DHCP springs to life and obtains a different address from my router.  This should then simply continue working going forward, until it doesn't.  

Why is This Happening

I have Googled this and never really found a straight answer, so I guess no one really knows.  It could be a bug in Windows, the router or some sort of bug between Parallels and macOS.  Suspicion probably falls more towards the router because, for example, I never see this issue if the MacBook is tethered to my Pixel 4a 5G.  

I will get around to tcpdump-ing and inspecting what happens in Wireshark, when time permits (but don't expect to learn anything useful).  

Caveats and Problems

Caveat works for me, obviously.

In terms of problems, when I first encountered this on the Mac, very recently, I initially considered just leaving the virtual machine with a fixed IP address.  However, I then noticed a bug whereby the Mac (which uses DHCP) would power up or wake up with the virtual machine's IP address (not in the router's DHCP range), causing a clash if I then fired up the Windows virtual machine.

The final caveat is I encountered this issue on an M1 MacBook using an Insider Build of Windows 10 for ARM.  So all fairly bleeding edge.  This might simply never be a problem on production builds running on Intel based Mac systems?

© Jason Hindle

Comments

Post a Comment

Popular posts from this blog

Thoughts on Face Masks

(They get in the way) Based on my recent comment to an article in The Register . The Argument For Face Masks? One of the surprising aspects of the pandemic is that a behaviour considered pretty normal, in much of the Far East and South Asia, is so controversial in West, attracting many conspiracy theories.  That act is wearing a face mask, in public, either where one has cold or flu, or during an epidemic, is done by pretty much anyone who lives in China, Hong Kong, Japan and South Korea.  In the Far East, personal responsibility is still very much a thing; people tend to keep their diseases to themselves and the masks are pretty ubiquitous at any time. So why do we wear face masks when in confinement? Masks are just used as things that get in the way of other things. In this case, we are hoping masks get in the way of airborne droplets containing a virus. Masks are porous, which is generally a good thing.  The wearer of a none porous mask would have a worse problem than ...
Post a Comment
Read more

The Office

(In which I ponder its future) A solitary visit to the office, early June left me in a somewhat sombre mood.  We've had the current office for over three years and, by any measure, this is a good office.  Once it was clear we had to make the move form our rather run down old office, a great deal of effort went into locating the new office and then designing the form and function of the place.  Not perfect and certainly not funky but nevertheless a pleasant, modern and productive place where, unlike the old office (pictured below in its sad, final weeks), I'm not the snot monster several days in a month. I can't speak for my employer's thinking, but what once looked like temporary changes, brought about by the pandemic, now has me contemplating the future of office spaces and work in general.  Since March we've worked from home with no apparent difference in productivity.  Will the old order re-assert itself, post pandemic? Of course, there is both a human and ec...
Post a Comment
Read more