New Mystery Issue - Machine Won't Boot or Get Past Initial Black Screen

[britechguy]

This is an issue I have never encountered before, and never encountered anything like it before, so I thought I'd ask if the symptoms are familiar to others.

The machine under discussion is the Dell Optiplex 3010 in which I replaced the PSA last week. This went perfectly, the machine booted back up just fine, and the client has been using it at home for days now. It was running Windows 10, Version 1903.

He says that last night he got a Windows Update notice that he needed to restart, so he did. The machine appears to have been doing for him what I will soon describe, but he did mention, "So I turned it off." I doubt this could have been a firmware flash gone wrong just based on what I could find today.

The Dell splash screen always shows.

When that closes, a black screen with the following text appears (and I may have missed some):

Intel Base-Code PXE-2.1 (Build 083)

Realtek PCIe GBP Family Controller v2.58 10/08/13

DHCP: (Twirling pipe character that eventually stops, after which the Client MAC Address & GUID are presented)

PXE-E53 No boot filename received

PXE-MOF Exiting PXE ROM

No boot device found

I ran Dell's On-Board Diagnostics and everything except Hard Drive passes, and for Hard Drive it presents:

Msg: Hard Drive - No Hard Drive detected or disk controller not supported.

-------------------------------------------

I've already unseated and reseated all cables, power and SATA, to the HDD and Mobo.

The obvious suspicion is that I have had a HDD failure, but this seems to be too obvious, not to mention a bit to coincidental for my liking. So I thought I'd ask if this sort of situation rings a bell for anyone, particularly with regard to the obvious fault possibly being reported spuriously.

I have not, as yet, pulled the HDD and connected it as an external drive to another machine to check its state when doing so.

If there are any specific further diagnostics I should be doing/checking, that would be good to know, too.

Thanks in advance for any and all assistance offered.

 

[Kitten Kong]

Does sound drive related to me.

Remember, it's not a case of if a drive fails, but when.

 

[sapphirescales]

a bit to coincidental for my liking.

You're kidding me, right? Windows updates are freaking huge and take forever on a traditional hard drive. For casual computer users, Windows updates stress the hardware more than they ever do browsing the internet. Hardware is more likely to fail when it's under stress and less likely to fail when it's under light load. In this case, "stress" is relative. If I had a dollar for every time a drive partially/completely failed while doing a Windows update, I'd be a very rich man.

 

[Porthos]

Might have switched from MBR to UEFI or vice versa.
I would also test the drive.

 

[Mike McCall]

Check the BIOS settings. It's looking for PXE, which is network boot. Might find something there.

 

[britechguy]

I intend to pull the drive and test it, but I'll see if I can pull the data from it first.

Like so many home users, no backups exist.

The system has secure boot off and is set to boot in legacy mode, so I'm presuming MBR format. Why this was done on a machine that has UEFI I'll never know. It also seems to me to be way overdue for a firmware update given that what's on it dates from 2013.

 

[Computer Bloke]

I intend to pull the drive and test it, but I'll see if I can pull the data from it first.

If you can spare thirty seconds, slap a known-good HDD in first and see if it shows up in the BIOS - if it does then you have a badly failed drive on your hands and will probably need specialist help.

I really wouldn't try to recover data before doing that, for all the obvious and well-known reasons. If by any chance they're not obvious and well known, here's the definitive guide: https://www.technibble.com/forums/resources/how-to-triage-a-hard-drive.17/

And (hoping this doesn't sound too patronizing) always remember that computers are just glorified light bulbs - they rarely fail in continuous use and usually go pop when turned off and on.

 

[sapphirescales]

Why this was done on a machine that has UEFI I'll never know.

Early UEFI was buggy. I personally would have just updated the BIOS to a newer, less buggy version, but another technician might not have wanted to take the risk.. Since it's an Optiplex, they usually get BIOS updates for at least 5 years. But some technicians subscribe to the old adage "if it ain't broke, don't fix it" and would rather just disable UEFI rather than update the BIOS.

Both trains of thought are valid. I personally would have updated the BIOS and installed Windows in UEFI mode, but 15 years ago, a younger me would have just installed in legacy mode (and yes, I'm aware that UEFI didn't exist back then, but if it had I would have just disabled it on an older system like this).

 

[britechguy]

If you can spare thirty seconds, slap a known-good HDD in first and see if it shows up in the BIOS - if it does then you have a badly failed drive on your hands and will probably need specialist help.

I really wouldn't try to recover data before doing that, for all the obvious and well-known reasons. If by any chance they're not obvious and well known, here's the definitive guide: https://www.technibble.com/forums/resources/how-to-triage-a-hard-drive.17/

And (hoping this doesn't sound too patronizing) always remember that computers are just glorified light bulbs - they rarely fail in continuous use and usually go pop when turned off and on.

None of the above "sounds patronizing" to me in any way. It is so very easy to overlook the obvious, and particularly if one is frustrated (which I am).

Having someone calmly remind me about things I should be taking into account is appreciated.

Also, I have never written responses strictly with the original questioner in mind. Those without as much experience frequently find topics like this via web search, and putting bits like this in place for future readers is doing a huge service.

As to the light-bulb analogy, you're right, and it's one of the reasons I don't frequently turn off my machines at all. They're on 24/7 the vast majority of the time.

 

[Markverhyden]

After confirming the failure, first thing I'd have done is boot from partedmagic. Then go from there.

 

[Sky-Knight]

Drive... be... bad...

When you pull it, you'll confirm it.

They always die on a restart... it's just the way it happens.

 

[britechguy]

They always die on a restart... it's just the way it happens.

<Voce de Church Lady>How Conveeeeeeeeeenient!</Voce de Church Lady>

And a big thanks to all who've responded.

 

[Sky-Knight]

@britechguy You did just replace the PSU, when it died did you test it?

Perhaps you had a dirty 5v rail on it? That'll cook drives, usually kills them with the PSU but it's possible for them to go later.

Hard to know for certain... at this point the fix is the same, replace the drive. The rest is just curiosity.

 

[britechguy]

Just replaced, which is what I always do. And since this was a Dell OEM PSU, I'd never have thought about testing to begin with. The old one was "dead, dead, dead," so there was nothing I could check on it.

I just said elsewhere, about an entirely different issue, I tend to concern myself with reasonably probable issues and not remotely possible ones. A bad 5v rail on a brand new PSU is remotely possible, at best, and since the drive has been functioning normally for days now I really doubt it's the issue.

And none of the above should be taken as dismissive or critical, just me thinking out loud.

 

[fincoder]

Windows updates are freaking huge and take forever on a traditional hard drive. For casual computer users, Windows updates stress the hardware more than they ever do browsing the internet. Hardware is more likely to fail when it's under stress and less likely to fail when it's under light load.

This, and the fact that HDDs fail so often. Coincidences involving HDDs are common. Also the drive may have been failing for some months prior, if so the timing in this case is unremarkable. Before quoting on a repair for an older system with HDD, I always run Crystal Disk Info to check the drive health because more often that not the HDD is also failing. If you knew it had faulty PSU and HDD it might have been better to replace the system.

The system has secure boot off and is set to boot in legacy mode, so I'm presuming MBR format. Why this was done on a machine that has UEFI I'll never know.

It dates from the early days of UEFI. PCs back then that came with Windows 7 default to legacy BIOS, the UEFI settings were for future use with Windows 8.

 

[Porthos]

Those without as much experience frequently find topics like this via web search

Not in this section. No sees it unless you are a member who has meet the post count and time delay, Not even Google.

 

[britechguy]

This, and the fact that HDDs fail so often. Coincidences involving HDDs are common.

I must live a very, very charmed life. I can count on less than 2 hands the number of failed HDDs I've had to deal with over the course of decades now.

When I first started in this business in the 1980s, HDD failure was common, relatively speaking. But, my heavens!, they've become incredibly more durable in recent decades.

That's one of the reasons this event threw me, I've never had an HDD fail under this circumstance, and out of the blue. There are generally lots of warning signs, and that start out slowly and can go on for a very long time, before an HDD fails.

 

[britechguy]

Not in this section. No sees it unless you are a member who has meet the post count and time delay, Not even Google.

But you can see it via the forums search, so . . .

 

[Porthos]

But you can see it via the forums search, so . . .

The point is it is private to eligible members only and will not show in any outside search.

Now that being said not everyone here is an actually experienced tech. Many are new to repairing computers and some slipped past the checks.

Back on topic, Some computers either during a firmware/bios update or even a low/bad bios battery will revert back to defaults of the board which might be different from the layout of the current install(mbr/uefi)

 

[Computer Bloke]

I can count on less than 2 hands the number of failed HDDs I've had to deal with over the course of decades now.

Either you've been very lucky or you're not looking hard enough. I can't even count on fewer than two hands the number of failed or failing HDDs I've seen this year.