Windows 7 update stuck on constant "need to restart" - end of my tether

[Carol Haynes]

I have spent most of the day on this and have tried everything I can think of - anyone got any other suggestions other than a lump hammer ...

Windows 7 did some updates the other day - now I am getting pop up reminders to restart

After numerous restarts the pops keep coming

I can't check for updates (I need to restart first).

I have run the Windows Update Trouble shooter numerous time --- comes up with an error unfixed every time but when I look at the detailed version it says it is fixed

Windows 7 update stuck on constant "need to restart" - end of my tether

Windows 7 update stuck on constant "need to restart" - end of my tether

I have run the MS FixIt Tool for Windows Update in non-aggressive and agressive mode --- says it is complete --- restart and back to square one. No no update history though.

I have manually reset Windows Update (renamed SoftwareDistribution and CatRoot2) --- makes no difference

I have deleted the registry key HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\WindowsUpdate\Auto Update\RebootRequired  --- it keeps reappearing and forcing the restart prompt

I have run SFC /SCANNOW - there are errors but they are not related directly to windows and have been there for ages without issue (they relate to 3rd party drivers)

I have restarted about 20 times through all this day and spent most of the day swearing when I should be doing something more useful.

Apart from wiping Windows and reinstalling anyone got any suggestions?

Thanks

[Arizona Hot]

Call Microsoft and scream when connected. Seriously, can you backup Windows to before the update. Windows should have set a restore point(I think) before updating. If you are connecting when the other computer has this problem, you can get Microsoft online and ask them what to do. Then you can blame them if they can't fix it.

[Carol Haynes]

I have followed all the advice from Microsoft - which ends with reinstall windows which I haven't got time to do at the moment.

I only have 6 system restore points - all from today.

[Arizona Hot]

Do you have backup software files you could use to restore the computer to a more manageable condition?

[cyberdiva]

Carol, I'm a little surprised you don't have a restore point from when the updates were installed. It's my impression that the installation of Windows updates almost always starts with Windows creating a restore point, if not for all the updates (if they're installed at more or less the same time), then at least for the first. Are you sure there aren't some additional restore points aside from those made today? (Could the dates be wrong and the restore points are not from today?)

Also, did the problems start right after the updates were installed, or did they start today? If by any chance the problems started today, the first of the restore points might possibly be useful.

I've always considered myself needlessly cautious in setting a restore point before I install Windows Updates. I knew (or at least I thought I knew) that Windows would create one, but my mistrust of MS usually won out.  Perhaps excessive caution with MS is never needless. 

Good luck in solving this. I wish I could help.

[x16wda]

Weird. I am sure you have looked at a hundred of the posts with "solutions". This computer is not in a domain, is it? Did you try creating a new administrator account and using that account to start the process?

You could try an in place upgrade to the same o/s, might be easier than a fresh Windows reinstall and all it takes is time.

[KodeZwerg]

Since no one asked yet (i do first that before anything else)
What does Windows Event Log tell when running that Update/Booting Pc ?
If you do not know Event Log, Microsoft TechNet teach it.

[4wd]

Apart from wiping Windows and reinstalling anyone got any suggestions?

-Carol Haynes (July 17, 2018, 12:14 PM)

Uninstall any AV/AM, do a partition backup, perform an In-Place Upgrade - no need to wipe Windows.  Takes an hour or so and all your files/programs are still installed - anything with some kind of virtual driver will possibly need reinstalling, (eg. DaemonTools).

EDIT: Just noticed x16wda suggested it, even so would have been my first option in the event of no pre-update backup.

[Carol Haynes]

Thanks for all the suggestions ... no it is not in a domain

No I don't have a useful backup as my stupid 3Tb Seagate drives packed up (one of the heads stopped working and a second drive purchased at the same time also had a failing head). I will never EVER buy Seagate crap again! I have been burned before but never again.

I left it updating over night last week and didn't immediately notice there was an issue until it started popping up saying I need to restart - I did but no sign of updates installing and when it rebooted it just popped up again which is when I checked what was going on. There are no older restore points.

A number of side issues - I have Windows 7 and 10 dual booting (mostly use 7 as I hate 10) but it has the unfortunate effect of I can't boot from a Windows 7 repair or installation DVD because the boot sectors tell it I have windows 10 and the installer complains it is not the right version. This means an in place upgrade is not happening - there doesn't seem to be any way of restoring the W7 boot code as I can't get to the 'Repair your computer screen' (I don't mind trashing the W10 dual boot if I can get the W7 boot system working again).

Re. system restore points - it does create system restore points - I can only assume that in all my efforts so many restore points have been created that the useful ones have been deleted.

I have been checking system integrity - SFC reports issues it can;t fix but there is noting the CBS.LOG file to give me a clue.

I ran Microsoft's System Readiness tool (which basically actsa as DSIM in Windows 7) - this was much more helpful and identified 8 corrupt or missing files. It fixed 7 of them. The 8th is listed as this:

   servicing\packages\Package_for_KB976933_vpc~31bf3856ad364e35~amd64~nb-NO~7.1.7601.17514.mum
   servicing\packages\Package_for_KB976933_vpc~31bf3856ad364e35~amd64~nb-NO~7.1.7601.17514.cat

The manual fix instructions say to download the correct from from MS Dowload Centre but nothing is listed there under KB976933

Download the package from Microsoft Download Center or Microsoft Update Catalog.
Copy the package (.msu) to the %SYSTEMROOT%\CheckSUR\packages directory. By default, this directory doesn't exist and you need to create the directory.
Rerun the System Update Readiness Tool.

I think the original file was part of SP1 - any ideas how to extract it from the SP1 package or do I need to reapply SP1 (if windows will let me - I seem to remember last time I tried that it just said already installed?)



System error log seems to have a lot of errors that look like this:

- System

  - Provider

   [ Name]  Microsoft-Windows-Kernel-Tm
   [ Guid]  {4CEC9C95-A65F-4591-B5C4-30100E51D870}
 
   EventID 1
 
   Version 0
 
   Level 3
 
   Task 0
 
   Opcode 0
 
   Keywords 0x8000000000000000
 
  - TimeCreated

   [ SystemTime]  2018-07-19T03:53:34.581684300Z
 
   EventRecordID 2332315
 
   Correlation
 
  - Execution

   [ ProcessID]  13864
   [ ThreadID]  21048
 
   Channel System
 
   Computer OFFICE-PC
 
  - Security

   [ UserID]  S-1-5-18
 

- EventData

  TxUow {DFDD7667-8AA1-11E8-99CA-50E549C546E8}
  TxDescriptionLength 0
  TxDescription 
  ClfsStatus 0xc0000061


The Transaction (UOW={dfdd7667-8aa1-11e8-99ca-50e549c546e8}, Description='') was unable to be committed, and instead rolled back; this was due to an error message returned by CLFS while attempting to write a Prepare or Commit record for the Transaction.  The CLFS error returned was: 0xc0000061.

The details vary but "was unable to be committed, and instead rolled back; this was due to an error message returned by CLFS while attempting to write a Prepare or Commit record for the Transaction.  The CLFS error returned was:" is common

There also seem to be quite a few of these:

- System

  - Provider

   [ Name]  Microsoft-Windows-BitLocker-Driver
   [ Guid]  {651DF93B-5053-4D1E-94C5-F6E6D25908D0}
 
   EventID 24620
 
   Version 0
 
   Level 2
 
   Task 0
 
   Opcode 0
 
   Keywords 0x8000000000000000
 
  - TimeCreated

   [ SystemTime]  2018-07-18T15:48:43.551730900Z
 
   EventRecordID 2330167
 
   Correlation
 
  - Execution

   [ ProcessID]  4
   [ ThreadID]  56
 
   Channel System
 
   Computer OFFICE-PC
 
  - Security

   [ UserID]  S-1-5-18
 

- EventData

  ErrorCode 0x80000010
  Volume \\?\Volume{3ccee3f6-1e28-4872-994b-a3c9066ba7e6}
  WritePhase 0x0

Encrypted volume check: Volume information on \\?\Volume{3ccee3f6-1e28-4872-994b-a3c9066ba7e6} cannot be read.

Which is odd (I am not using BitLocker - the drives are not encrypted). I suppose this might be related to drives that have been removed (the knackered Seagate drives for a start).


I am away for 10 days from tomorrow - I am not being rude if I don't reply to ideas quickly - I will be slaving behind a cello and my only tech access will be sporadic via an Android phone. I will check out any ideas when I get home. Thanks everyone.

[Shades]

The Hirens BootCD project is under new management. Well, it is now maintained by volunteers who only use freeware/open-source tools (and some trial versions) for repair from lots and lots of different types of errors that Windows subjects its users to.

You can make a bootable pen drive with the iso they provide. To be prepared for calamities, I created myself such a pendrive and it boots very fast into a Windows 10 PE (x64) environment. Afterwards the 16GByte pendrive was filled with 6 to 7 GByte from my personal PortableApps collection of tools. And I guess about 70% of the added tools works as well. The ones that don't, report that prerequisites are missing, which is to be expected from a PE environment.

Booting takes about 20 to 25 seconds (my pendrive is a USB3 model, but it is slow as hell) when connected to an USB3 port on an APU-based AMD (A10-6800K) system with 8GByte of RAM. Even on a nothing fancy PC, it boots up fast. If push comes to shove, it could even be good enough as a daily driver from the get-go, especially when you just want to browse a bit.

The included repair tools are quite impressive, but it can also do backup and forensic tasks as well. And it is more than likely this PE environment can help you fix your Windows 7 issues.

On another note:
While the interface of Windows 10 is not everyone's cup of tea, it is not that bad in my opinion. Having said that, Google reveals quite a lot of links that can make your Windows 10 installation look and "feel" like Windows 7. In that sense you would have the best of 2 worlds and you could get rid of dual-boot altogether. From Gizmo's Freeware or find your favorite link with this search term: make Windows 10 look and feel like Windows 7

[Carol Haynes]

On another note:
While the interface of Windows 10 is not everyone's cup of tea, it is not that bad in my opinion. Having said that, Google reveals quite a lot of links that can make your Windows 10 installation look and "feel" like Windows 7. In that sense you would have the best of 2 worlds and you could get rid of dual-boot altogether. From Gizmo's Freeware or find your favorite link with this search term: make Windows 10 look and feel like Windows 7

-Shades (July 21, 2018, 01:57 AM)

Trouble is I have some stuff I need that does not work well on Windows 10 so need Windows 7 at the moment.

Its not really a look and feel thing (I can live with W10 style) but more importantly the thing that winds me up about Windows 10 are the constant updates that break things - the last major feature update (Fall Creator Edition) left my system completely unbootable (including Windows 7) and it took me days to get anything working again and involved reinstalling Windows 10 from clean as it wouldn't even start recovery from USB or CD. I haven't let it install the Spring Creators Update and don't trust it too but I can't stop it forever so I am thinking of simply dumping Windows 10 on this machine. I have fixed too many customer computers recently that have had problems with the Spring update (two of them needed wiping and starting again).

I just bought a 1Tb Samsung SSD and graphics card and plan to unplug all the existing drives and do a clean W7 install on the SSD and then connect by dual booting with my exisiting Windows 7 until I can transfer everything and make sure it works. That is a long term project though - and to be honest I am not sure all the software is still installable (eg. I have an old copy of Quickbooks 2012 and don;t need to pay a monthly subscription to pay for features I will never use - just not sure I can get that activated and all the updates - if not there is no support for recovering data from my QB file so I may just need to clean up the existing install and make it minimal to use odd bits like this).

[IainB]

...No I don't have a useful backup as my stupid 3Tb Seagate drives packed up (one of the heads stopped working and a second drive purchased at the same time also had a failing head). I will never EVER buy Seagate crap again! I have been burned before but never again. ...

-Carol Haynes (July 20, 2018, 08:35 PM)

I know it's pedantic, but it seems to me that it could be useful here to note three points which could stand in the way of learning as much as one might be able to from this discussion:

• Objectivity: A hard drive being simply a manmade electro-mechanical device, it would be incorrect and an anthropomorphism (the attribution of human characteristics or behaviour to a god, animal, or object) to call it "stupid".
• Statistical validity: Without further substantive data on relevant hard drive failure-rates and performance, it would be incorrect to attribute statistical significance to one person's anecdote/experience of eventual failure of only two drives of the same brand.
• Logical fallacy: The fallacy of the availability heuristic may also be in operation (prediction of the frequency of an event, or a proportion within a population, based on how easily an example can be brought to mind) - e.g., "One swallow does not a summer make".

The engineers who design and manufacture hard drives would be well aware of both the statistical validity and relevance of the data on their manufacturing processes (they would generally be processes in statistical control) and of the performance characteristics of their hard drive equipment in testing and in operation out in the field and would have amassed much statistical performance data on their hard drive componentry and on that of their competitors.
The ordinary user can avail themselves of some this data from their HDDs today, which accumulate/log (in on-board counters) self-sensed/reported data - the S.M.A.R.T. variables - and which can be analysed to provide performance reports for the HDD in question - e.g., using various software tools including (say) SpeedFan or (better) Hard Drive Sentinel.

Left undisturbed, operational HDDs don't usually "suddenly fail", they tend to progressively fail in incremental fashion, all the meanwhile logging their gradual deterioration in their on-board SMART data accumulators. Analysis of HDD failure can be quite enlightening and can enable the user to predict HDD failure, when they avail themselves of SMART monitoring to detect when it is time to migrate from a failing drive, rather than risk blindly waiting till forced to recover from an already-failed drive.

Some potentially relevant/useful links on DC Forum:

• (a) Brand Reliability: Hard Drive Brand Reliability Data (also see other references in the DC Forum to Backblaze analysis).
  
  
• (b) Monitoring Hard Drive deterioration:Hard Disk Sentinel PRO - Mini-Review
  
  
• (c) Using SMART analysis to predict failure: Re: Hard Drive SMART Stats - from the BackBlaze Blog.
  
  
• (d) Ease of migrating (for a novice) in planned fashion from a known failing HDD:
  

  I downloaded and installed the smaller file, then executed AOMEI Backupper. Click-click-click with the mouse - and in less than 20 seconds I had started cloning my failing Seagate drive to the new WD drive.
  It was by then Saturday 2AM. I sat with it for a couple of hours, interested in seeing its progress reports, and making sure that the laptop would not go to sleep and set the screen to switch off after 2 minutes of no keyboard activity (so as to keep the temps down inside the laptop). I estimated that at the rate it was going it would take about 8 hours to complete, and so I went to bed. I was awoken at about 8AM by the HDS alarm gong telling me that the health of the Seagate drive had deteriorated another 1% to 54%, and I noticed that AOMEI Backupper was still busy making the clone with apparently no problem.
  

  -IainB (November 23, 2013, 05:51 AM)

[Carol Haynes]

Statistical validity: Without further substantive data on relevant hard drive failure-rates and performance, it would be incorrect to attribute statistical significance to one person's anecdote/experience of eventual failure of only two drives of the same brand.

-IainB (July 21, 2018, 02:38 PM)

The drives I bought were subject of a large class action lawsuit because they were notoriously unreliable and gave no warning of sudden failure.

Its not the first time seagate drives I bought had this reputation  within months of product launch.

Whether my experience is statistically relevant I no longer trust Seagate drives.

[f0dder]

Left undisturbed, operational HDDs don't usually "suddenly fail", they tend to progressively fail in incremental fashion, all the meanwhile logging their gradual deterioration in their on-board SMART data accumulators. Analysis of HDD failure can be quite enlightening and can enable the user to predict HDD failure, when they avail themselves of SMART monitoring to detect when it is time to migrate from a failing drive, rather than risk blindly waiting till forced to recover from an already-failed drive.

-IainB (July 21, 2018, 02:38 PM)

That's not the experience I've had as a consumer.

In some cases, I've seen "reallocated sector count" or some similar stat go up before an eventual full breakdown - but mostly, it's been "worked fine yesterday, now I can't get my data". And I've had disks with reallocater sectors that kept trucking along for years without flaw, and just ended up being too small.

SMART is a mess. The values are opaque, and you can't really compare them between brands. There's no guarantee you'll get reported errors before a failure, and reported errors are no guarante of a failure. And, moving from spinning magnetic platters to solid state drives, failures tend to be "oops, logic board died, all data is lost".

[IainB]

@Carol Haynes:
Wow! Did Seagate lose to the "large class action lawsuit" (that would be in the US, I guess)?
I did write: "Without further substantive data on relevant hard drive failure-rates and performance,...", but presumably there would have been substantive data aplenty to back up a class action lawsuit - no?

Double-Wow!: "Those drives "were notoriously unreliable and gave no warning of sudden failure" - see, that's what I meant by "process in statistical control". If Seagate production processes had been operated as being "in statistical control", then such glaring defects as to cause failure without warning (i.e., no SMART problem alarm triggers) couldn't have got out the door - à la Deming/Shewhart.
From experience, I have enormous respect for the ability of design engineers of this kind of technology. I reckon they could never have allowed such garbage to get out the door and be shipped to customers, thus I would presume that management probably screwed up big-time - e.g., maybe time-to-market had to be short to generate quick revenues, so they had to save time and money by short-cutting on quality control, or something.
Shambolic. A recipe for corporate suicide.

Your experience, coupled with the rest of the members of the class action would surely have to be statistically relevant  - if they won the class action.

"I no longer trust Seagate drives" - I don't blame you. Neither would I! Would have left a very bad taste in the mouth.

Ah, just found it here: <https://www.hbsslaw.com/cases/seagate>
(Copied below sans embedded hyperlinks/images; some emphasis is mine.)

Seagate Hard Drives
DEFENDANT NAME: Seagate Technologies LLC
CASE NUMBER: 5:16-cv-00523
COURT: U.S. District Court for the Northern District of California
PRACTICE AREA: Consumer Rights
STATUS: Active
DATE FILED: 02/01/16

KEY ATTORNEYS: Steve W. Berman
                          Jeff D. Friedman

RELATED DOCUMENTS:
Order 02/09/17
Consolidated Complaint 05/09/16
Complaint 02/05/16
Complaint 02/01/16
CONTACT: [email protected]

COURT ISSUES ORDER, UPHOLDS CLAIMS
Judge Spero of the Northern District of California ruled on Seagate’s motion to dismiss on Feb. 9, 2017. In so doing, he refused to dismiss (1) plaintiffs’ consumer protection claims based on Seagate’s statements as to the drives’ annualized failure rates and suitability for use in certain RAID configurations, and (2) plaintiffs’ unjust enrichment claims. Plaintiffs will continue to pursue these claims vigorously. Read the order »

Did you buy Seagate’s Barracuda 3TB Hard Disk Drive, Backup Plus 3TB External Hard Disk Drive or another Seagate hard drive with model number ST3000DM001? You may be entitled to reimbursement under consumer laws.
FILL OUT THE FORM TO FIND OUT YOUR RIGHTS »
Hagens Berman and Axler Goldich LLC are investigating Seagate Technologies for selling hard drives that routinely failed at exceptionally high rates, leaving consumers with broken hardware and significant loss of data. Consumers filed a national class-action lawsuit, and the law firms are seeking more information from affected consumers.

These particular hard drives were marketed as innovative, fast, powerful, reliable, dependable, and having extremely low failure rates, when in reality, the failure rate of the drives was substantially higher than advertised. Consumers report them failing as an unprecedented rate – sometimes even days after their first use.

According to the firms' investigation, Seagate promised purchasers that it would replace the failed hard drives, but replacements were also defective, and failed at extremely high rates, leaving Seagate’s warranty promise unfulfilled, and consumers without working hard drives.

Consumers have reportedly lost tons of data unexpectedly, as Seagate’s hard drives failed to live up to the advertised promises, violating consumer laws.

If you purchased Seagate’s Barracuda 3TB Hard Disk Drive, Desktop HDD 3TB, Backup Plus 3TB External Hard Disk Drive, GoFlex 3TB External Hard Disk Drive, or another Seagate hard drive with model number ST3000DM001, you may be entitled to damages including replacement costs and damages from loss of data and data recovery expenses.
Note: the internal model is called Seagate Barracuda or Seagate 3TB Desktop HDD, but it still has the same model number as the Barracuda.

Find out your rights. Fill out the form to contact our legal team.

Very interesting.

[wraith808]

Left undisturbed, operational HDDs don't usually "suddenly fail", they tend to progressively fail in incremental fashion, all the meanwhile logging their gradual deterioration in their on-board SMART data accumulators. Analysis of HDD failure can be quite enlightening and can enable the user to predict HDD failure, when they avail themselves of SMART monitoring to detect when it is time to migrate from a failing drive, rather than risk blindly waiting till forced to recover from an already-failed drive.

-IainB (July 21, 2018, 02:38 PM)

That's not the experience I've had as a consumer.

In some cases, I've seen "reallocated sector count" or some similar stat go up before an eventual full breakdown - but mostly, it's been "worked fine yesterday, now I can't get my data". And I've had disks with reallocater sectors that kept trucking along for years without flaw, and just ended up being too small.

SMART is a mess. The values are opaque, and you can't really compare them between brands. There's no guarantee you'll get reported errors before a failure, and reported errors are no guarante of a failure. And, moving from spinning magnetic platters to solid state drives, failures tend to be "oops, logic board died, all data is lost".

-f0dder (July 21, 2018, 05:55 PM)

Totally agreed with all of the above.  I've had hard drives fail with varying symptoms or no symptoms, running diagnostic software that said that there were no flaws on the drive until it actually failed- then all of a sudden the scan was showing a lot of errors right as it failed using HDD Sentinel.

[KodeZwerg]

If you ask 10 people wich Manufactor of Harddrives is bad you have at least 5-7 different choices.
My 10000 rpm Seagate Barracuda (SCSI) is more than 10 Years old without any issues, Western Digital Drives are the bad ones for me.

To your Windows Problem:
Why dont you just run Windows 10 within VirtualMachine? This way you can easy (if such feature is activated) roll-back to any prior state.
This works way better than, excuse me, dumb restore points that to 90% does not do what you try to do.
If update/software you installed in VM seems to be okay you can start thinking about to make it real.
I configured my VM to never save anything unless i want to save changes.

[xtabber]

Seagate 3TB drives were notoriously prone to failure (I had one, and it did, after about 3 years).

According to the latest BackBlaze survey, the more recent Seagate 4TB drives have lower failure rates than either HGST or WD.

[Carol Haynes]

For anyone interested I found a fix for this issue - at an admin command prompt:


fsutil resource setautoreset true c:\

Then in SAFE MODE or Recovery Console at command prompt:

C: (in Recovery console use the drive indicated)
cd Windows\System32\SMI\Store\Machine
attrib -s -h *
del *.blf
del *.regtrans-ms
cd Windows\System32\config\TxR
attrib -s -h *
del /q *



Restart!! Fixed!! Finally!!! Hope someone else finds it useful