Jump to content


Photo

Main PC took 6 hours - 8 attempts to boot...

Backing up rapidly - about to

  • Please log in to reply
17 replies to this topic

#1 g__day

g__day

    Champion

  • Hero
  • 7,885 posts

Posted 07 November 2017 - 03:51 PM

WOnder if anyone has ever experienced this - my main PC BSOD this morning - Kernal Exception - wouldn't reboot clean despite multiple attempts, safe mode wouldn't work, Auto system repair hung... basically total freeze.

 

Rebooted off Win 10 Install DVD - hung saying a driver was needed - WTF?  Tried again and after 4 hours with the WIndows logo and the six spinning white dots it let me in normally - and I am now backing up files like crazy!

 

What does it do that takes hours to get to the Windows password screen with no status or warning?  Why wouldn't safe mode boot work nor System repair?  Memry check was all I could do - it worked fine.

 

Now my search for a new PC is on steroids - will porbably go down the Intel I7 8720X route and do a start from scratch build 32 GB RAM, 1100W PSU 1080 ti as main video card with a 500 GB 960 Pro as the main boot drive and add one or two large HDDs and transfer all the files I wish to keep across.

 

Glad at least I could log in today!

 

PS WOndering if I should try a windows restore or not - hate it if it BSOD half way thru...  Note when I tried to run CPUID today - it only gets 80% thru the load - halting on Storage services check!


Edited by g__day, 07 November 2017 - 03:57 PM.

Talent + Integrity = Atomic!

#2 Rybags

Rybags

    Immortal

  • Super Hero
  • 35,058 posts

Posted 07 November 2017 - 04:05 PM

My guess - PSU and/or motherboard on the way out.  Or maybe dodgy Ram.

I'd be going minimal system and try and id the faulty component by elimination.

 

ed - or maybe just some faulty card or peripheral.  But same methodology applies for working it out.


Edited by Rybags, 07 November 2017 - 04:06 PM.


#3 g__day

g__day

    Champion

  • Hero
  • 7,885 posts

Posted 07 November 2017 - 04:22 PM

Too many parts mate - 8 HHDs 4 SSDs so many add on parts and cards, PSU is a Corsair 850W job - be surprised if that has given up the ghost.  CPU / MB is an OLD Conroe2 Q6600 on a Gigabyte P35 motherboard.  It has done the job well for many, many years - so I think it desrved a well earned rest!

 

Reckon I will spend $4K - $5K and buy something decent, all pre made by someone who knows that they are doing.  Any recommendations about good stores in Sydney to do the job plus check the specs look sensible / well matched?  I haven't ordered a high end build in ages as I usually get gear form a local or go Umart, PCCase gear, AusPCmarket.

 

Could folk here recommend a really good PC shop - high parts and build quality?

 

Many thanks!


Talent + Integrity = Atomic!

#4 Rybags

Rybags

    Immortal

  • Super Hero
  • 35,058 posts

Posted 07 November 2017 - 04:58 PM

Do a quick visual of the board, swollen caps and shit-stains near components esp mosfets around the CPU.

 

P35 board... so probably mid 2007 thru 2008 or so.  I've got a P45 board here of theirs, from what I've experienced still seems to work OK, has a waterblock setup so I've flogged the arse out of a few CPUs on it.


Edited by Rybags, 07 November 2017 - 04:58 PM.


#5 g__day

g__day

    Champion

  • Hero
  • 7,885 posts

Posted 07 November 2017 - 05:39 PM

I am thinking its all just too old (like me) and had one too many refreshes.  When you're not sure if a WIndows refresh will make things better or worse = wow!


Talent + Integrity = Atomic!

#6 Rybags

Rybags

    Immortal

  • Super Hero
  • 35,058 posts

Posted 07 November 2017 - 05:42 PM

Win10 installs are anything but immune to being corrupted.

Sometimes a fresh install is what's needed.  Of course in the modern day that's a huge PITA - 8 minutes to get the install then 8 hours for drivers, updates and getting all your programs reinstalled.



#7 g__day

g__day

    Champion

  • Hero
  • 7,885 posts

Posted 07 November 2017 - 06:27 PM

My worry is it dying in the middle due to flakey hardware - making it all the more high risk either way.


Talent + Integrity = Atomic!

#8 Rybags

Rybags

    Immortal

  • Super Hero
  • 35,058 posts

Posted 07 November 2017 - 06:36 PM

What I'd do since it's so old is unhook the drives and do a Win install to a blank drive, then have it running burn-in type software overnight... something like the old Sandra test that flog the memory and give the CPU a workout too.



#9 Nich...

Nich...

    Professional Tart

  • Mod
  • 43,271 posts
  • Location:Mexico

Posted 07 November 2017 - 08:28 PM

Yeah. If you're not sure what's flakey, back it up by putting the drives in a different box.
"I think it is a sad reflection on our civilization that while we can and do measure the temperature in the atmosphere of Venus we do not know what goes on inside our soufflés" -- Nicholas Kurti

#10 g__day

g__day

    Champion

  • Hero
  • 7,885 posts

Posted 07 November 2017 - 09:02 PM

Just re-booted - then noted (after 90 minutes) when it got thru the spinning white dots - in Task Manager -> Performance -> Disk 8 C: General Stuff was 100% active for over 30 minutes, whilst every other drive - even the Windows 10 drive is only 0% - 1% active.

 

So I may try to find and remove this drive and see what happens.  8 HDDs - 3 are 3 TB Samsungs like this one - only three turns at most to find the culprit!


Talent + Integrity = Atomic!

#11 Rybags

Rybags

    Immortal

  • Super Hero
  • 35,058 posts

Posted 07 November 2017 - 09:17 PM

Check Event Viewer logs too... the trick is to set a filter so you're not getting the information msgs, only the error ones.  If a drive is failing it'll usually generate some entries.



#12 g__day

g__day

    Champion

  • Hero
  • 7,885 posts

Posted 08 November 2017 - 11:35 AM

Didn't see any entries there related to disks - but I just powered off my B, C and D drives - and things are flying again at the moment and boot times are back to normal.  PC seems much more responsive.

 

I ponder that some of the WIndows Storage integrity check software may be single threaded and it hangs everything if a storage device doesn't respond in time or hangs occassionally returning status.  Bit of a novice mistake when coding Operating Systems - you presume tasks may fail and you set a timeout limit and complain gracefully when they fail and if you deem them non essential you don't crap out with a BSOD!


Talent + Integrity = Atomic!

#13 merlin13

merlin13

    Guru

  • Atomican
  • 8,402 posts

Posted 11 November 2017 - 06:20 PM

...but I just powered off my B, C and D drives - and things are flying again at the moment and boot times are back to normal.  PC seems much more responsive...

Haaaaang on a minute - B drive? B??... for an actual hard disk?

 

Could be just me but who else would raise an eyebrow at that?


Anything not nailed down is mine, and anything I can prise up is not nailed down...

#14 Rybags

Rybags

    Immortal

  • Super Hero
  • 35,058 posts

Posted 11 November 2017 - 07:05 PM

In theory you could rename normal drives to use those letters.  ATM I use "O:" for DVD (as in optical) and use Y/Z for virtual drives (image mounts).



#15 merlin13

merlin13

    Guru

  • Atomican
  • 8,402 posts

Posted 12 November 2017 - 11:07 AM

Yeah, the option may be there to do that but historically A: 'n B: has/was used for floppies - using one of those for a solid drive nowadays on a WinDoze 10 system is just a bit kinda... asking for the onset of Ye Goode Olde  Undocumented Feature, y'know?... Something I wouldn't be brave enough to try with a drive full of stuff I couldn't afford to lose and/or have to dick about reloading.

 

Ah, well - as the OP's boot problem after shutting those drives down obviously points towards a flaky drive, flaky controller/cable etc, or even a new Undocumented WinDoze 10 Feature perhaps (smarm smarm pending), guess we'll have to wait and see exactly what went wrong.


Anything not nailed down is mine, and anything I can prise up is not nailed down...

#16 g__day

g__day

    Champion

  • Hero
  • 7,885 posts

Posted 12 November 2017 - 12:22 PM

I do recall when this PC actually still had floppies in those drive bays!  I even recall when I took out the last 40 pin 80 wire drive connector to old HDDs and CD drive.  Yes when I installed my 3rd 3TB drive I thought I would use it for Backup - so when letters from A-Z - minus those already taken turned up I choose B for backup!

 

I truly do wonder why one failing I/O device of 12 - and a fairly unimportnat, irregularily used one - cause Windows 10 so much angst!  Why it can't just say -uh oh - looks like a failing drive - device 10 B: or whatever is beyond me.  If there is an external USB3 - SATA drive bay - might get one and test these drives on several PCs.  It could be drive controllers, cable, motherboard, power cables etc.PC still has crashed several times - Kernal State exceptions and Defered Procedure Calls - so something is still wrong.  Have run major virus scans, Registery checkers and optimisers, driver up to date checkers etc.  So once I am on the new PC with my data across I think I will simplify this machine then do a full W10 re-install.

 

I look forward to having a rock solid PC again in a week or two!


Edited by g__day, 19 November 2017 - 10:05 AM.

Talent + Integrity = Atomic!

#17 Nich...

Nich...

    Professional Tart

  • Mod
  • 43,271 posts
  • Location:Mexico

Posted 12 November 2017 - 01:29 PM

I've had an 8TB drive that's been problematic over the last year or so.  Normally it works fine, but every few months, windows has a hard time detecting it as connected.  When that happens, i/o on every drive grinds to a halt as windows has a panic attack trying to work out why it can't see that drive that should be there and is still getting power.

 

It's set up as a removable drive (all sata drives are) via ahci or whatever, but it's the unexpected nature of it appearing then disapearing that seems to cause problems.


/anecdote


"I think it is a sad reflection on our civilization that while we can and do measure the temperature in the atmosphere of Venus we do not know what goes on inside our soufflés" -- Nicholas Kurti

#18 Master_Scythe

Master_Scythe

    Titan

  • Hero
  • 20,181 posts
  • Location:QLD

Posted 13 November 2017 - 09:22 AM

I've had an 8TB drive that's been problematic over the last year or so.  Normally it works fine, but every few months, windows has a hard time detecting it as connected.  When that happens, i/o on every drive grinds to a halt as windows has a panic attack trying to work out why it can't see that drive that should be there and is still getting power.

 

It's set up as a removable drive (all sata drives are) via ahci or whatever, but it's the unexpected nature of it appearing then disapearing that seems to cause problems.


/anecdote

 

This is one place that any form of *nix is a legend.

 

I don't know enough about Kernels to know why, but all I know is that things like hanging USB sticks, faulty HDD's, even 'forever spooling printers' can cause the whole OS to absolutely become a typewriter; although less useful because the typewriter is giving output :P

 

On a *nix machine, you can get a single thread\daemon hung, but it never seems to break the rest of the OS.


Wherever you go in life, watch out for Scythe, the tackling IT support guy.

"I don't care what race you are, not one f*cking bit, if you want to be seen as a good people, you go in there and you f*ck up the people who (unofficially) represent you in a negative light!"





0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users