I would also add that Ubuntu failed to boot up fully on 3 or 4 occasions as well, adding to the concern that using the ports on the left would not be totally reliable.

Same thing happened to me, and I also made the connection to the USB port having a plugged-in drive. No issues on Windows, though I had Ubuntu installed on the internal hard disk.

@FRAPONTILLO

Thanks for commenting.

So you had Ubuntu installed on the hard drive?

Did the installed Ubuntu restart when shut down as well or just the USB plugged-in drive?

Oh, yeah... I have a USB scanner (not too old - Fujitsu ScanSnap S1100 ) that won't talk to the Dell. Any similar issues?

I have the same issue. Using Fedora on an internal SSD. Even on battery power. I have Win install media on order to test on that OS. I doubt it helus. It even happens on shutdown from GRUB an BIOS setup (via the power button. Ugh.

Interim solution: be ready to push the power button as fast as possible  if you see the keyboard light up. And if at first you don't succeed: push, push again.

Thanks for your input, MIKEHAT.NET.

I had forgotten about the power button till I read your post.

On the occasions that the laptop restarted after shutting down Ubuntu from the USB, I was usually letting it boot up again to test a few more times.

But on some occasions where I wanted to stop it with the power button, it would even restart again from there.

While I have read elsewhere that there may be a bug in the kernel causing this, I don't see what the kernel has to do with a situation such as this.

I have now reluctantly returned my laptop but will continue to monitor threads for any potential solutions.

Better luck with the new machine? Mine requires the power button game every time now, but I put Win 8 on an internal drive and it powers off correctly. Seems like it's a linux kernel problem to me...

I'd like to hear that a new machine has different results. I will use it with linux no matter what the problem is.

Thanks once again for your input.

Sadly, I took a refund for the machine I returned and bought a second hand ultrabook that had popped up instead (not Dell). I thought I'd try the smaller screen format for a while, though I did want the 15.6".

It has no such 'restarting' issues with Ubuntu though and it almost all worked 'out of the box'.

Here's what worked:     sudo sync. Just one time fixed it for good (20 shutdowns and counting).

Came across this post while searching for a similar issue. I had recently installed Ubuntu 14.04 on my Dell Inspiron 7537. Shutdowns were not performing correctly (i.e. acting as a reboot rather than a shutdown). Neither command line execution of shutdown nor shutdown option from the gear icon in the menu bar worked, only holding down the power button for a while seemed to work. 

I ran sudo sync and it fixed the issue so far. Will repost if anything reverts to incorrect behavior again. Does anyone understand why this solution works?

Edit: I was too quick to post...come to find out sudo sync didn't fix it but I started to come across what I believe may be the issue. The shutdown sequence seems work properly when I have a device (specifically a USB mouse) plugged into a USB port (specifically the lower left port). Without the USB device in the port, the system performs a reboot rather than a shutdown. This leads me to believe that there may be a setting in the BIOS which I can modify. There are several USB settings which pertain to USB options that I'm going to test out. I've came across other posts with similar shutdown/reboot issues on other platforms where Ethernet Wake on LAN settings in the BIOS seemed to be causing the issue which leads me to believe this could be something similar.

Hello, I have same problem with same model. Dell inspiron 15 7000. 

Initially with win8.1were resets when i try to shutdown. But not always. Maybe once for 10 shutdowns.

Then I try to install ubuntu to see if this happens.

After installation doesn't happens, but after full online updates of ubuntu, then happens every time!!!

I install ubuntu without win8.1

Then i have installed a lot of linux distributions to see if this happens. And yes!!! I have installed: ubuntu, fedora, opensuse, Mint, debian, Manjaro. Yes i am lazy, but i wanted to make sure. 

With all distributions resets every time I shutdown. 

With win8.1 also happens but not everytime. As i sayd before 1\10 times.

Also I tried to update Bios to 06 version. And reset defaults.

The problem persists.

Is there any solution?

I reported earlier that I had solved the problem. But getting the same issue, and have noticed the same results with plugging external drives into the usb ports.

BTW, I'm using a SSD as the internal drive. I wonder if others with this issue are doing the same.

Also, I don't believe mine is actually doing a reboot. A reboot happens immediately after the shutdown, but this restart issue takes about 5 seconds to get going. And I think the solution lies in having the kernel wait a bit before turning off the hardware. I think this because if I boot to the grub2 command line, wait for a little while and then issue the halt command, the machine shuts down reliably.

Mine is a Dell Inspiron 15 7537.

256GB SSD 16GB RAM.

Does restarts instead of shutdown, with o without devices conected in USB ports.

Restarts happens 3-5 seconds after keyboard lights turn off.

Is not only in Linux, also happens in Windows 8.1 but less.

In Linux, after new installation, is working ok. In my case the problem starts after a full online update of any linux distribution.

Im very frustrated with this Dell.

Well, OlderNWiser has moved on to pastures. Read the whole discussion, as it's taken on a life of its own. This is not a power supply issue. It is a hardware/firmware issue related to shutting down the board.

I think that with an external USB device plugged in, either the system waits long enough for the devicce to shut down or the power draw of the device helps quiet things down. It is possible that some new-age disk caching feature is keeping the system awake after the kernel calls it quits.

BTW, I am using Fedora, not Ubuntu. Win8 has the same (albeit less reproducible) problem.

Regarding the base (non-updated) installs of some OSes working, perhaps it is a set of delayed and/or last-second disk writing that triggers the problem. I doubt that the updates are bringing in kernel updates that cause this, as I noticed this even after doing a fresh install with the newer kernel - no restart problem until updates are applied.

The sync & sudo sync resolution is a semi-solution, so I will try to get my system to issue that at the tail end of the shutdown sequence and see what happens...

Post edited to acknowledge Canonical/Dell arrangement.

Hi OlderNWiser,

     That's quite a post - lots of good detail. Here are a couple things that might open this issue up to further discussion...although if you've already scheduled return of your machine it may be a moot point.

     First. I'm following that you suspect dirty power. I'm also wondering why this is your conclusion after reading your post twice. What are the common symptoms you typically see with a noisy supply in other devices? The "brick" of your charger/ac adapter is built with redundant ferrous core filters inline with supply and output and ac/dc conversion is achieved by waveform inversion. Luckily with your background in electronics you should be able to put it on a scope and check output for things like fuzz or spikes or drops...any deviation from the listed output voltage and amperage. It should be measured under an imposed load. Doing so may actually help identify a larger problem if it is in fact poor supply conditioning which would benefit other Forum Members with similar problems.

     Dell does support UBUNTU and for the GNU-Linux savvy UBUNTU should function as well as the next OS. I personally installed UBUNTU on a machine that had a Turion x64 processor only to find after successful installation that the machine would hang during boot and generate a looping error that forced me to hard start the machine each time I wanted to change or adjust something. This was in October 2013 and after several frustrating days I learned from the UBUNTU support forum that There was a conflict with how the Turion X64 processes instructions....never to be resolved since that line is obsolete. I find it likely that with newer bigger faster instruction sets being prevalent in current processors it could be a bug that UBUNTU hasn't ironed out yet.

      I almost never (let me emphasize almost) suspect a crippling hardware problem that only effects one installation, whether it's an OS or an upgrade, a driver or an application. These sort of problems tend to follow specific lines of function. For example, UBUNTU wouldn't be more electrically sensitive that say Windows 8.1Pro x64. This is especially true for basic, primitive devices like usb ports and bus. (5 volt core by any other name is a 5 volt core right?) the rest is simply I/O that comes along for the ride....which just now prompts me to consider; you may collect additional power quality data by also scoping the USB+ and USB- at each of the ports with the unit both plugged in as well as on battery power. Compare them to known working conterparts as a control.

     I would at some point like to give UBUNTU another shot but it probably won't happen until my next laptop purchase. In the meantime I'm really interested in your findings and hope you still have your machine long enough to get some of the data. Looking forward to your next visit to the Forum!