I use Windows Live Mail 11 (WLM) on this Windows 7 Home Premium machine.
Lately, I've been getting an error message when I try to load WLM that
tells me the Calendar is corrupt, error 0x8E5E0247. Here's a screenshot:

https://postimg.cc/kVSv7zX5

Winows 7 Home Premium, SP1
Windows Live Essentials 2011, installed 9/21/2011
Live Mail, defailt email program, use Thunderbird when needed
Live Messenger (never used, don't have a MSFT account)
Windows Live Photo Gallery, I use it, works fine
Windows Live Movie Maker, I use it, works fine

The hard drive on this machine is a 1TB SATA Seagate ST31000528AS, 70% free

Years ago, I saved "wlsetup-all.exe", 220,149 KB. When WLM began to fail a
few days ago with the 0x8E5E0247 error, on this Windows 7 machine, I
installed wlsetup-all.exe on my Windows 10 Home Premium desktop, deleted
the WLM storage file that was created upon initial install, and replaced
the storage file with the storage file from my Windows 7 WLM. The Windows
10 WLM picked up all of the emails/folders, etc that were on the Windows 7
WLM machine. Great. I can now use the Windows 10 desktop to collect
email, and it has all my old emails, too. But, I'd like to get the Windows
7 WLM working again, because it's the machine I sit at the most, in the
room I am in the most.

By the way, WLM has also been working successfully on my Windows 10 Home
Premium, fully updated, laptop.

From everything I've tried and read about the 0x8E5E0247, I think the
problem may be an Intel Storage Driver/Manager/Controller.
(Deleting/remaning the Calendar folder has no effect.) For instance, I
read, "...uninstalling the storage manager (e.g. Intel Rapid Storage
Manager or Intel Matrix Storage Manager, the software will vary with
drive/system manufacturer) and then installing the latest version."

My device manager shows this driver, and when I try to update, I'm told I
have the best driver around.

https://postimg.cc/yDJS4G7r

My Control Panel shows this:

https://postimg.cc/0KSzSRF6

Any suggestions as next steps? I haven't tried deleteing any Intell
'stuff', since I don't know if I could get them back. But, I do have a
backup. Any search and destroy changes to the registry recommended?

TIA

On 7/18/2022 10:29 PM, Boris wrote:
> I use Windows Live Mail 11 (WLM) on this Windows 7 Home Premium machine.
> Lately, I've been getting an error message when I try to load WLM that
> tells me the Calendar is corrupt, error 0x8E5E0247. Here's a screenshot:
>
> https://postimg.cc/kVSv7zX5
>
> Winows 7 Home Premium, SP1
> Windows Live Essentials 2011, installed 9/21/2011
> Live Mail, defailt email program, use Thunderbird when needed
> Live Messenger (never used, don't have a MSFT account)
> Windows Live Photo Gallery, I use it, works fine
> Windows Live Movie Maker, I use it, works fine
>
> The hard drive on this machine is a 1TB SATA Seagate ST31000528AS, 70% free
>
> Years ago, I saved "wlsetup-all.exe", 220,149 KB. When WLM began to fail a
> few days ago with the 0x8E5E0247 error, on this Windows 7 machine, I
> installed wlsetup-all.exe on my Windows 10 Home Premium desktop, deleted
> the WLM storage file that was created upon initial install, and replaced
> the storage file with the storage file from my Windows 7 WLM. The Windows
> 10 WLM picked up all of the emails/folders, etc that were on the Windows 7
> WLM machine. Great. I can now use the Windows 10 desktop to collect
> email, and it has all my old emails, too. But, I'd like to get the Windows
> 7 WLM working again, because it's the machine I sit at the most, in the
> room I am in the most.
>
> By the way, WLM has also been working successfully on my Windows 10 Home
> Premium, fully updated, laptop.
>
> From everything I've tried and read about the 0x8E5E0247, I think the
> problem may be an Intel Storage Driver/Manager/Controller.
> (Deleting/remaning the Calendar folder has no effect.) For instance, I
> read, "...uninstalling the storage manager (e.g. Intel Rapid Storage
> Manager or Intel Matrix Storage Manager, the software will vary with
> drive/system manufacturer) and then installing the latest version."
>
> My device manager shows this driver, and when I try to update, I'm told I
> have the best driver around.
>
> https://postimg.cc/yDJS4G7r
>
> My Control Panel shows this:
>
> https://postimg.cc/0KSzSRF6
>
> Any suggestions as next steps? I haven't tried deleteing any Intell
> 'stuff', since I don't know if I could get them back. But, I do have a
> backup. Any search and destroy changes to the registry recommended?
>
> TIA
>

There is a theory here, in the last posting.

https://answers.microsoft.com/en-us/windowslive/forum/all/windows-live-mail-calendar-contains-corrupt-data/4e4eb047-c603-480c-aae4-67798cada343

"Today I applied a Windows patch that was supposed to fix a problem
that was forecast to occur with Live Mail from 2016. That appeared
to stop the program crashing."

"I followed your advice and renamed the file. Mail now seems stable.
I was even able to open the Calendar and add an appointment."

%Localappdata%\Microsoft\Windows Live Mail\Calendars ???

I think you've already tried that though.

The Intel driver is "RAID-ready". That means when the driver
first installs and all your disks are JBOD, the driver operates
in AHCI mode. If you flip to RAID mode, any drive that did not
specifically get added to a RAID array, is still in JBOD.

Only if you had setup RAID, would the driver be doing RAID. Like
striped or mirrored or RAID5 parity.

This scheme is designed to allow "RAID Migration". Have a single
drive with C: on it. Add a blank drive to machine. Make an array
from original C: and blank drive, and it automatically converts
it to RAID 0/1/5 and so on. This was intended to solve the problem
which existed previously, where it was so hard to get a RAID install going.

If the Intel driver was fouled up, why would the boot even finish ?
If you only had a single hard drive, the hard drive would be
running as AHCI (using the AHCI file in the Intel RAID folder).

*******

OMG. The stupid stuff you find on teh Internet.

https://wlmail.wordpress.com/esent-errors/

0x8E5E0247 SectorSizeNotSupported The physical sector size reported
by the disk subsystem, is unsupported
by ESE for a specific file type.

That would be stuff like 512e (emulation), 512n (native), 4Kn and so on.

512e 4096 physical 512 virtual Only an issue with WinXP alignment, OK on Win7 alignment
512n 512 physical 512 virtual Bulletproof compatibility, traditional size, WDC Gold small disks
4Kn 4096 physical 4096 virtual Do not buy, use, or mix these with desktop PCs... Trouble awaits!
May need to buy different/new tools, for stuff.

512e is a lot more common than 512n today, and while some brands may have 512n,
it could disappear in any model year. One of the drives in the Win7 machine
is a 512n, and suited to seamless usage in WinXP (which can no longer happen here).

These have been around long enough, this is surely some kind of
bogus error.

https://mskb.pkisolutions.com/kb/2470478

"ESENT expects that the reported physical sector size remains *consistent*
between sessions. When the ESENT database re-initializes, it makes sure
that the physical sector size that the database was created for and the
current reported physical sector size are the same. When the sector sizes
are different, ESENT reports an error."

The ST31000528AS appears to be 512n. The spec is dated 2009, Wiki says
advanced format drives were around 2010. That's why the spec does not
contain the nomenclature expected.

As administrator:

fsutil fsinfo ntfsinfo c:

NTFS Volume Serial Number : 0x1234abcd1234abcd
NTFS Version : 3.1
Bytes Per Sector : 512 <=== LogicalSector \__ a "512 emulated"
Bytes Per Physical Sector : 4096 <=== PhysicalSector / drive
Bytes Per Cluster : 4096 (4 KB) file system storage unit
Bytes Per FileRecord Segment : 1024 <=== $MFT slot size

Check and see what info you can find on yours. Yours is currently
likely to be 512 512 4096 1024. If you were in RAID mode (two drives),
I don't know what they stuff in those fields for that.

*******

If the config is working in Windows 10, including calendar, could
you bring the materials back to the Win7 machine ?

Winston seems to know a lot of trivia for WLM, perhaps he will
have a suggestion.

But on the surface, this is just... ridiculous. Absurd. Like a steam powered rocket.
This can't be it, can it?

Paul

On 7/19/2022 3:05 AM, Paul wrote:
> On 7/18/2022 10:29 PM, Boris wrote:
>> I use Windows Live Mail 11 (WLM) on this Windows 7 Home Premium machine.
>> Lately, I've been getting an error message when I try to load WLM that
>> tells me the Calendar is corrupt, error 0x8E5E0247.  Here's a
>> screenshot:
>>
>> https://postimg.cc/kVSv7zX5
>>
>> Winows 7 Home Premium, SP1
>> Windows Live Essentials 2011, installed 9/21/2011
>>          Live Mail, defailt email program, use Thunderbird when needed
>>          Live Messenger (never used, don't have a MSFT account)
>>          Windows Live Photo Gallery, I use it, works fine
>>          Windows Live Movie Maker, I use it, works fine
>>
>> The hard drive on this machine is a 1TB SATA Seagate ST31000528AS,
>> 70% free
>>
>> Years ago, I saved "wlsetup-all.exe", 220,149 KB.  When WLM began to
>> fail a
>> few days ago with the 0x8E5E0247 error, on this Windows 7 machine, I
>> installed wlsetup-all.exe on my Windows 10 Home Premium desktop, deleted
>> the WLM storage file that was created upon initial install, and replaced
>> the storage file with the storage file from my Windows 7 WLM. The
>> Windows
>> 10 WLM picked up all of the emails/folders, etc that were on the
>> Windows 7
>> WLM machine.  Great.  I can now use the Windows 10 desktop to collect
>> email, and it has all my old emails, too.  But, I'd like to get the
>> Windows
>> 7 WLM working again, because it's the machine I sit at the most, in the
>> room I am in the most.
>>
>> By the way, WLM has also been working successfully on my Windows 10 Home
>> Premium, fully updated, laptop.
>>
>>  From everything I've tried and read about the 0x8E5E0247, I think the
>> problem may be an Intel Storage Driver/Manager/Controller.
>> (Deleting/remaning the Calendar folder has no effect.) For instance, I
>> read, "...uninstalling the storage manager (e.g. Intel Rapid Storage
>> Manager or Intel Matrix Storage Manager, the software will vary with
>> drive/system manufacturer) and then installing the latest version."
>>
>> My device manager shows this driver, and when I try to update, I'm
>> told I
>> have the best driver around.
>>
>> https://postimg.cc/yDJS4G7r
>>
>> My Control Panel shows this:
>>
>> https://postimg.cc/0KSzSRF6
>>
>> Any suggestions as next steps?  I haven't tried deleteing any Intell
>> 'stuff', since I don't know if I could get them back.  But, I do have a
>> backup.  Any search and destroy changes to the registry recommended?
>>
>> TIA
>>
>
> There is a theory here, in the last posting.
>
> https://answers.microsoft.com/en-us/windowslive/forum/all/windows-live-mail-calendar-contains-corrupt-data/4e4eb047-c603-480c-aae4-67798cada343
>
>
>    "Today I applied a Windows patch that was supposed to fix a problem
>     that was forecast to occur with Live Mail from 2016. That appeared
>     to stop the program crashing."
>
>    "I followed your advice and renamed the file. Mail now seems stable.
>     I was even able to open the Calendar and add an appointment."
>
>     %Localappdata%\Microsoft\Windows Live Mail\Calendars ???
>
> I think you've already tried that though.
>
> The Intel driver is "RAID-ready". That means when the driver
> first installs and all your disks are JBOD, the driver operates
> in AHCI mode. If you flip to RAID mode, any drive that did not
> specifically get added to a RAID array, is still in JBOD.
>
> Only if you had setup RAID, would the driver be doing RAID. Like
> striped or mirrored or RAID5 parity.
>
> This scheme is designed to allow "RAID Migration". Have a single
> drive with C: on it. Add a blank drive to machine. Make an array
> from original C: and blank drive, and it automatically converts
> it to RAID 0/1/5 and so on. This was intended to solve the problem
> which existed previously, where it was so hard to get a RAID install
> going.
>
> If the Intel driver was fouled up, why would the boot even finish ?
> If you only had a single hard drive, the hard drive would be
> running as AHCI (using the AHCI file in the Intel RAID folder).
>
> *******
>
> OMG. The stupid stuff you find on teh Internet.
>
> https://wlmail.wordpress.com/esent-errors/
>
>    0x8E5E0247 SectorSizeNotSupported  The physical sector size reported
>                                       by the disk subsystem, is
> unsupported
>                                       by ESE for a specific file type.
>
> That would be stuff like 512e (emulation), 512n (native), 4Kn and so on.
>
>    512e  4096 physical  512 virtual    Only an issue with WinXP
> alignment, OK on Win7 alignment
>    512n   512 physical  512 virtual    Bulletproof compatibility,
> traditional size, WDC Gold small disks
>    4Kn   4096 physical 4096 virtual    Do not buy, use, or mix these
> with desktop PCs... Trouble awaits!
>                                        May need to buy different/new
> tools, for stuff.
>
> 512e is a lot more common than 512n today, and while some brands may
> have 512n,
> it could disappear in any model year. One of the drives in the Win7
> machine
> is a 512n, and suited to seamless usage in WinXP (which can no longer
> happen here).
>
> These have been around long enough, this is surely some kind of
> bogus error.
>
> https://mskb.pkisolutions.com/kb/2470478
>
>    "ESENT expects that the reported physical sector size remains
> *consistent*
>     between sessions. When the ESENT database re-initializes, it makes
> sure
>     that the physical sector size that the database was created for
> and the
>     current reported physical sector size are the same. When the
> sector sizes
>     are different, ESENT reports an error."
>
> The ST31000528AS appears to be 512n. The spec is dated 2009, Wiki says
> advanced format drives were around 2010. That's why the spec does not
> contain the nomenclature expected.
>
>    As administrator:
>
>    fsutil fsinfo ntfsinfo c:
>
>    NTFS Volume Serial Number :        0x1234abcd1234abcd
>    NTFS Version      :                3.1
>    Bytes Per Sector  :                512            <===
> LogicalSector  \__ a "512 emulated"
>    Bytes Per Physical Sector :        4096           <===
> PhysicalSector /   drive
>    Bytes Per Cluster :                4096  (4 KB)        file system
> storage unit
>    Bytes Per FileRecord Segment    :  1024           <=== $MFT slot size
>
> Check and see what info you can find on yours. Yours is currently
> likely to be 512 512 4096 1024. If you were in RAID mode (two drives),
> I don't know what they stuff in those fields for that.
>
> *******
>
> If the config is working in Windows 10, including calendar, could
> you bring the materials back to the Win7 machine ?
>
> Winston seems to know a lot of trivia for WLM, perhaps he will
> have a suggestion.
>
> But on the surface, this is just... ridiculous. Absurd. Like a steam
> powered rocket.
> This can't be it, can it?
>
>    Paul
>
You got me thinking about the drive set up on the Windows 7 machine. 
There are two SATA drives attached as RAID:

1) 1TB SATA Seagate ST31000528AS, my C:, with OS, programs, data, etc.

2) 1TB SATA WDC10EZEX-08WN4A0, my F:, a clone of C: from when C: was
reinstalled fresh with Windows 7 and programs, including WLM, about
three years ago;

I tried booting from 2), and launching WLM.  No luck, same error.

I disconnected power from 2) and machine booted from 1).  WLM launched
just fine.  I've shut down WLM many times, and it comes back up fine.  I
guess next step is to re-attach 2) and see if the error returns when WLM
is launched from 1).

If it's 2) that's suddenly causing the error, it may explain why a
restored image of an 'error free' backup STILL gave the error.

On 7/19/2022 3:03 PM, Boris wrote:
> On 7/19/2022 3:05 AM, Paul wrote:
>> On 7/18/2022 10:29 PM, Boris wrote:
>>> I use Windows Live Mail 11 (WLM) on this Windows 7 Home Premium
>>> machine.
>>> Lately, I've been getting an error message when I try to load WLM that
>>> tells me the Calendar is corrupt, error 0x8E5E0247.  Here's a
>>> screenshot:
>>>
>>> https://postimg.cc/kVSv7zX5
>>>
>>> Winows 7 Home Premium, SP1
>>> Windows Live Essentials 2011, installed 9/21/2011
>>>          Live Mail, defailt email program, use Thunderbird when needed
>>>          Live Messenger (never used, don't have a MSFT account)
>>>          Windows Live Photo Gallery, I use it, works fine
>>>          Windows Live Movie Maker, I use it, works fine
>>>
>>> The hard drive on this machine is a 1TB SATA Seagate ST31000528AS,
>>> 70% free
>>>
>>> Years ago, I saved "wlsetup-all.exe", 220,149 KB.  When WLM began to
>>> fail a
>>> few days ago with the 0x8E5E0247 error, on this Windows 7 machine, I
>>> installed wlsetup-all.exe on my Windows 10 Home Premium desktop,
>>> deleted
>>> the WLM storage file that was created upon initial install, and
>>> replaced
>>> the storage file with the storage file from my Windows 7 WLM. The
>>> Windows
>>> 10 WLM picked up all of the emails/folders, etc that were on the
>>> Windows 7
>>> WLM machine.  Great.  I can now use the Windows 10 desktop to collect
>>> email, and it has all my old emails, too.  But, I'd like to get the
>>> Windows
>>> 7 WLM working again, because it's the machine I sit at the most, in the
>>> room I am in the most.
>>>
>>> By the way, WLM has also been working successfully on my Windows 10
>>> Home
>>> Premium, fully updated, laptop.
>>>
>>>  From everything I've tried and read about the 0x8E5E0247, I think the
>>> problem may be an Intel Storage Driver/Manager/Controller.
>>> (Deleting/remaning the Calendar folder has no effect.) For instance, I
>>> read, "...uninstalling the storage manager (e.g. Intel Rapid Storage
>>> Manager or Intel Matrix Storage Manager, the software will vary with
>>> drive/system manufacturer) and then installing the latest version."
>>>
>>> My device manager shows this driver, and when I try to update, I'm
>>> told I
>>> have the best driver around.
>>>
>>> https://postimg.cc/yDJS4G7r
>>>
>>> My Control Panel shows this:
>>>
>>> https://postimg.cc/0KSzSRF6
>>>
>>> Any suggestions as next steps?  I haven't tried deleteing any Intell
>>> 'stuff', since I don't know if I could get them back.  But, I do have a
>>> backup.  Any search and destroy changes to the registry recommended?
>>>
>>> TIA
>>>
>>
>> There is a theory here, in the last posting.
>>
>> https://answers.microsoft.com/en-us/windowslive/forum/all/windows-live-mail-calendar-contains-corrupt-data/4e4eb047-c603-480c-aae4-67798cada343
>>
>>
>>    "Today I applied a Windows patch that was supposed to fix a problem
>>     that was forecast to occur with Live Mail from 2016. That appeared
>>     to stop the program crashing."
>>
>>    "I followed your advice and renamed the file. Mail now seems stable.
>>     I was even able to open the Calendar and add an appointment."
>>
>>     %Localappdata%\Microsoft\Windows Live Mail\Calendars ???
>>
>> I think you've already tried that though.
>>
>> The Intel driver is "RAID-ready". That means when the driver
>> first installs and all your disks are JBOD, the driver operates
>> in AHCI mode. If you flip to RAID mode, any drive that did not
>> specifically get added to a RAID array, is still in JBOD.
>>
>> Only if you had setup RAID, would the driver be doing RAID. Like
>> striped or mirrored or RAID5 parity.
>>
>> This scheme is designed to allow "RAID Migration". Have a single
>> drive with C: on it. Add a blank drive to machine. Make an array
>> from original C: and blank drive, and it automatically converts
>> it to RAID 0/1/5 and so on. This was intended to solve the problem
>> which existed previously, where it was so hard to get a RAID install
>> going.
>>
>> If the Intel driver was fouled up, why would the boot even finish ?
>> If you only had a single hard drive, the hard drive would be
>> running as AHCI (using the AHCI file in the Intel RAID folder).
>>
>> *******
>>
>> OMG. The stupid stuff you find on teh Internet.
>>
>> https://wlmail.wordpress.com/esent-errors/
>>
>>    0x8E5E0247 SectorSizeNotSupported  The physical sector size reported
>>                                       by the disk subsystem, is
>> unsupported
>>                                       by ESE for a specific file type.
>>
>> That would be stuff like 512e (emulation), 512n (native), 4Kn and so on.
>>
>>    512e  4096 physical  512 virtual    Only an issue with WinXP
>> alignment, OK on Win7 alignment
>>    512n   512 physical  512 virtual    Bulletproof compatibility,
>> traditional size, WDC Gold small disks
>>    4Kn   4096 physical 4096 virtual    Do not buy, use, or mix these
>> with desktop PCs... Trouble awaits!
>>                                        May need to buy different/new
>> tools, for stuff.
>>
>> 512e is a lot more common than 512n today, and while some brands may
>> have 512n,
>> it could disappear in any model year. One of the drives in the Win7
>> machine
>> is a 512n, and suited to seamless usage in WinXP (which can no longer
>> happen here).
>>
>> These have been around long enough, this is surely some kind of
>> bogus error.
>>
>> https://mskb.pkisolutions.com/kb/2470478
>>
>>    "ESENT expects that the reported physical sector size remains
>> *consistent*
>>     between sessions. When the ESENT database re-initializes, it
>> makes sure
>>     that the physical sector size that the database was created for
>> and the
>>     current reported physical sector size are the same. When the
>> sector sizes
>>     are different, ESENT reports an error."
>>
>> The ST31000528AS appears to be 512n. The spec is dated 2009, Wiki says
>> advanced format drives were around 2010. That's why the spec does not
>> contain the nomenclature expected.
>>
>>    As administrator:
>>
>>    fsutil fsinfo ntfsinfo c:
>>
>>    NTFS Volume Serial Number :        0x1234abcd1234abcd
>>    NTFS Version      :                3.1
>>    Bytes Per Sector  :                512            <===
>> LogicalSector  \__ a "512 emulated"
>>    Bytes Per Physical Sector :        4096           <===
>> PhysicalSector /   drive
>>    Bytes Per Cluster :                4096  (4 KB)        file system
>> storage unit
>>    Bytes Per FileRecord Segment    :  1024           <=== $MFT slot size
>>
>> Check and see what info you can find on yours. Yours is currently
>> likely to be 512 512 4096 1024. If you were in RAID mode (two drives),
>> I don't know what they stuff in those fields for that.
>>
>> *******
>>
>> If the config is working in Windows 10, including calendar, could
>> you bring the materials back to the Win7 machine ?
>>
>> Winston seems to know a lot of trivia for WLM, perhaps he will
>> have a suggestion.
>>
>> But on the surface, this is just... ridiculous. Absurd. Like a steam
>> powered rocket.
>> This can't be it, can it?
>>
>>    Paul
>>
> You got me thinking about the drive set up on the Windows 7 machine. 
> There are two SATA drives attached as RAID:
>
> 1) 1TB SATA Seagate ST31000528AS, my C:, with OS, programs, data, etc.
>
> 2) 1TB SATA WDC10EZEX-08WN4A0, my F:, a clone of C: from when C: was
> reinstalled fresh with Windows 7 and programs, including WLM, about
> three years ago;
>
> I tried booting from 2), and launching WLM.  No luck, same error.
>
> I disconnected power from 2) and machine booted from 1).  WLM launched
> just fine.  I've shut down WLM many times, and it comes back up fine. 
> I guess next step is to re-attach 2) and see if the error returns when
> WLM is launched from 1).
>
> If it's 2) that's suddenly causing the error, it may explain why a
> restored image of an 'error free' backup STILL gave the error.
>
>
I forgot to say that I've rebooted with only 1) connected, and WLM still
comes up fine.  In the morning, I'll re-attach 2), boot from 1), and see
if I get the WLM error message.

On 7/20/2022 12:16 AM, Boris wrote:
> On 7/19/2022 3:03 PM, Boris wrote:
>> On 7/19/2022 3:05 AM, Paul wrote:
>>> On 7/18/2022 10:29 PM, Boris wrote:
>>>> I use Windows Live Mail 11 (WLM) on this Windows 7 Home Premium machine.
>>>> Lately, I've been getting an error message when I try to load WLM that
>>>> tells me the Calendar is corrupt, error 0x8E5E0247.  Here's a screenshot:
>>>>
>>>> https://postimg.cc/kVSv7zX5
>>>>
>>>> Winows 7 Home Premium, SP1
>>>> Windows Live Essentials 2011, installed 9/21/2011
>>>>          Live Mail, defailt email program, use Thunderbird when needed
>>>>          Live Messenger (never used, don't have a MSFT account)
>>>>          Windows Live Photo Gallery, I use it, works fine
>>>>          Windows Live Movie Maker, I use it, works fine
>>>>
>>>> The hard drive on this machine is a 1TB SATA Seagate ST31000528AS, 70% free
>>>>
>>>> Years ago, I saved "wlsetup-all.exe", 220,149 KB.  When WLM began to fail a
>>>> few days ago with the 0x8E5E0247 error, on this Windows 7 machine, I
>>>> installed wlsetup-all.exe on my Windows 10 Home Premium desktop, deleted
>>>> the WLM storage file that was created upon initial install, and replaced
>>>> the storage file with the storage file from my Windows 7 WLM. The Windows
>>>> 10 WLM picked up all of the emails/folders, etc that were on the Windows 7
>>>> WLM machine.  Great.  I can now use the Windows 10 desktop to collect
>>>> email, and it has all my old emails, too.  But, I'd like to get the Windows
>>>> 7 WLM working again, because it's the machine I sit at the most, in the
>>>> room I am in the most.
>>>>
>>>> By the way, WLM has also been working successfully on my Windows 10 Home
>>>> Premium, fully updated, laptop.
>>>>
>>>>  From everything I've tried and read about the 0x8E5E0247, I think the
>>>> problem may be an Intel Storage Driver/Manager/Controller.
>>>> (Deleting/remaning the Calendar folder has no effect.) For instance, I
>>>> read, "...uninstalling the storage manager (e.g. Intel Rapid Storage
>>>> Manager or Intel Matrix Storage Manager, the software will vary with
>>>> drive/system manufacturer) and then installing the latest version."
>>>>
>>>> My device manager shows this driver, and when I try to update, I'm told I
>>>> have the best driver around.
>>>>
>>>> https://postimg.cc/yDJS4G7r
>>>>
>>>> My Control Panel shows this:
>>>>
>>>> https://postimg.cc/0KSzSRF6
>>>>
>>>> Any suggestions as next steps?  I haven't tried deleteing any Intell
>>>> 'stuff', since I don't know if I could get them back.  But, I do have a
>>>> backup.  Any search and destroy changes to the registry recommended?
>>>>
>>>> TIA
>>>>
>>>
>>> There is a theory here, in the last posting.
>>>
>>> https://answers.microsoft.com/en-us/windowslive/forum/all/windows-live-mail-calendar-contains-corrupt-data/4e4eb047-c603-480c-aae4-67798cada343
>>>
>>>    "Today I applied a Windows patch that was supposed to fix a problem
>>>     that was forecast to occur with Live Mail from 2016. That appeared
>>>     to stop the program crashing."
>>>
>>>    "I followed your advice and renamed the file. Mail now seems stable.
>>>     I was even able to open the Calendar and add an appointment."
>>>
>>>     %Localappdata%\Microsoft\Windows Live Mail\Calendars ???
>>>
>>> I think you've already tried that though.
>>>
>>> The Intel driver is "RAID-ready". That means when the driver
>>> first installs and all your disks are JBOD, the driver operates
>>> in AHCI mode. If you flip to RAID mode, any drive that did not
>>> specifically get added to a RAID array, is still in JBOD.
>>>
>>> Only if you had setup RAID, would the driver be doing RAID. Like
>>> striped or mirrored or RAID5 parity.
>>>
>>> This scheme is designed to allow "RAID Migration". Have a single
>>> drive with C: on it. Add a blank drive to machine. Make an array
>>> from original C: and blank drive, and it automatically converts
>>> it to RAID 0/1/5 and so on. This was intended to solve the problem
>>> which existed previously, where it was so hard to get a RAID install going.
>>>
>>> If the Intel driver was fouled up, why would the boot even finish ?
>>> If you only had a single hard drive, the hard drive would be
>>> running as AHCI (using the AHCI file in the Intel RAID folder).
>>>
>>> *******
>>>
>>> OMG. The stupid stuff you find on teh Internet.
>>>
>>> https://wlmail.wordpress.com/esent-errors/
>>>
>>>    0x8E5E0247 SectorSizeNotSupported  The physical sector size reported
>>>                                       by the disk subsystem, is unsupported
>>>                                       by ESE for a specific file type.
>>>
>>> That would be stuff like 512e (emulation), 512n (native), 4Kn and so on.
>>>
>>>    512e  4096 physical  512 virtual    Only an issue with WinXP alignment, OK on Win7 alignment
>>>    512n   512 physical  512 virtual    Bulletproof compatibility, traditional size, WDC Gold small disks
>>>    4Kn   4096 physical 4096 virtual    Do not buy, use, or mix these with desktop PCs... Trouble awaits!
>>>                                        May need to buy different/new tools, for stuff.
>>>
>>> 512e is a lot more common than 512n today, and while some brands may have 512n,
>>> it could disappear in any model year. One of the drives in the Win7 machine
>>> is a 512n, and suited to seamless usage in WinXP (which can no longer happen here).
>>>
>>> These have been around long enough, this is surely some kind of
>>> bogus error.
>>>
>>> https://mskb.pkisolutions.com/kb/2470478
>>>
>>>    "ESENT expects that the reported physical sector size remains *consistent*
>>>     between sessions. When the ESENT database re-initializes, it makes sure
>>>     that the physical sector size that the database was created for and the
>>>     current reported physical sector size are the same. When the sector sizes
>>>     are different, ESENT reports an error."
>>>
>>> The ST31000528AS appears to be 512n. The spec is dated 2009, Wiki says
>>> advanced format drives were around 2010. That's why the spec does not
>>> contain the nomenclature expected.
>>>
>>>    As administrator:
>>>
>>>    fsutil fsinfo ntfsinfo c:
>>>
>>>    NTFS Volume Serial Number :        0x1234abcd1234abcd
>>>    NTFS Version      :                3.1
>>>    Bytes Per Sector  :                512            <=== LogicalSector  \__ a "512 emulated"
>>>    Bytes Per Physical Sector :        4096           <=== PhysicalSector /   drive
>>>    Bytes Per Cluster :                4096  (4 KB)        file system storage unit
>>>    Bytes Per FileRecord Segment    :  1024           <=== $MFT slot size
>>>
>>> Check and see what info you can find on yours. Yours is currently
>>> likely to be 512 512 4096 1024. If you were in RAID mode (two drives),
>>> I don't know what they stuff in those fields for that.
>>>
>>> *******
>>>
>>> If the config is working in Windows 10, including calendar, could
>>> you bring the materials back to the Win7 machine ?
>>>
>>> Winston seems to know a lot of trivia for WLM, perhaps he will
>>> have a suggestion.
>>>
>>> But on the surface, this is just... ridiculous. Absurd. Like a steam powered rocket.
>>> This can't be it, can it?
>>>
>>>    Paul
>>>
>> You got me thinking about the drive set up on the Windows 7 machine. There are two SATA drives attached as RAID:
>>
>> 1) 1TB SATA Seagate ST31000528AS, my C:, with OS, programs, data, etc.
>>
>> 2) 1TB SATA WDC10EZEX-08WN4A0, my F:, a clone of C: from when C: was reinstalled fresh with Windows 7 and programs, including WLM, about three years ago;
>>
>> I tried booting from 2), and launching WLM.  No luck, same error.
>>
>> I disconnected power from 2) and machine booted from 1).  WLM launched just fine.  I've shut down WLM many times, and it comes back up fine. I guess next step is to re-attach 2) and see if the error returns when WLM is launched from 1).
>>
>> If it's 2) that's suddenly causing the error, it may explain why a restored image of an 'error free' backup STILL gave the error.
>>
>>
> I forgot to say that I've rebooted with only 1) connected, and WLM still comes up fine.  In the morning, I'll re-attach 2), boot from 1), and see if I get the WLM error message.

On 7/20/2022 3:14 AM, Paul wrote:
> On 7/20/2022 12:16 AM, Boris wrote:
>> On 7/19/2022 3:03 PM, Boris wrote:
>>> On 7/19/2022 3:05 AM, Paul wrote:
>>>> On 7/18/2022 10:29 PM, Boris wrote:
>>>>> I use Windows Live Mail 11 (WLM) on this Windows 7 Home Premium
>>>>> machine.
>>>>> Lately, I've been getting an error message when I try to load WLM
>>>>> that
>>>>> tells me the Calendar is corrupt, error 0x8E5E0247. Here's a
>>>>> screenshot:
>>>>>
>>>>> https://postimg.cc/kVSv7zX5
>>>>>
>>>>> Winows 7 Home Premium, SP1
>>>>> Windows Live Essentials 2011, installed 9/21/2011
>>>>>          Live Mail, defailt email program, use Thunderbird when
>>>>> needed
>>>>>          Live Messenger (never used, don't have a MSFT account)
>>>>>          Windows Live Photo Gallery, I use it, works fine
>>>>>          Windows Live Movie Maker, I use it, works fine
>>>>>
>>>>> The hard drive on this machine is a 1TB SATA Seagate ST31000528AS,
>>>>> 70% free
>>>>>
>>>>> Years ago, I saved "wlsetup-all.exe", 220,149 KB.  When WLM began
>>>>> to fail a
>>>>> few days ago with the 0x8E5E0247 error, on this Windows 7 machine, I
>>>>> installed wlsetup-all.exe on my Windows 10 Home Premium desktop,
>>>>> deleted
>>>>> the WLM storage file that was created upon initial install, and
>>>>> replaced
>>>>> the storage file with the storage file from my Windows 7 WLM. The
>>>>> Windows
>>>>> 10 WLM picked up all of the emails/folders, etc that were on the
>>>>> Windows 7
>>>>> WLM machine.  Great.  I can now use the Windows 10 desktop to collect
>>>>> email, and it has all my old emails, too.  But, I'd like to get
>>>>> the Windows
>>>>> 7 WLM working again, because it's the machine I sit at the most,
>>>>> in the
>>>>> room I am in the most.
>>>>>
>>>>> By the way, WLM has also been working successfully on my Windows
>>>>> 10 Home
>>>>> Premium, fully updated, laptop.
>>>>>
>>>>>  From everything I've tried and read about the 0x8E5E0247, I think
>>>>> the
>>>>> problem may be an Intel Storage Driver/Manager/Controller.
>>>>> (Deleting/remaning the Calendar folder has no effect.) For
>>>>> instance, I
>>>>> read, "...uninstalling the storage manager (e.g. Intel Rapid Storage
>>>>> Manager or Intel Matrix Storage Manager, the software will vary with
>>>>> drive/system manufacturer) and then installing the latest version."
>>>>>
>>>>> My device manager shows this driver, and when I try to update, I'm
>>>>> told I
>>>>> have the best driver around.
>>>>>
>>>>> https://postimg.cc/yDJS4G7r
>>>>>
>>>>> My Control Panel shows this:
>>>>>
>>>>> https://postimg.cc/0KSzSRF6
>>>>>
>>>>> Any suggestions as next steps?  I haven't tried deleteing any Intell
>>>>> 'stuff', since I don't know if I could get them back. But, I do
>>>>> have a
>>>>> backup.  Any search and destroy changes to the registry recommended?
>>>>>
>>>>> TIA
>>>>>
>>>>
>>>> There is a theory here, in the last posting.
>>>>
>>>> https://answers.microsoft.com/en-us/windowslive/forum/all/windows-live-mail-calendar-contains-corrupt-data/4e4eb047-c603-480c-aae4-67798cada343
>>>>
>>>>
>>>>    "Today I applied a Windows patch that was supposed to fix a problem
>>>>     that was forecast to occur with Live Mail from 2016. That appeared
>>>>     to stop the program crashing."
>>>>
>>>>    "I followed your advice and renamed the file. Mail now seems
>>>> stable.
>>>>     I was even able to open the Calendar and add an appointment."
>>>>
>>>>     %Localappdata%\Microsoft\Windows Live Mail\Calendars ???
>>>>
>>>> I think you've already tried that though.
>>>>
>>>> The Intel driver is "RAID-ready". That means when the driver
>>>> first installs and all your disks are JBOD, the driver operates
>>>> in AHCI mode. If you flip to RAID mode, any drive that did not
>>>> specifically get added to a RAID array, is still in JBOD.
>>>>
>>>> Only if you had setup RAID, would the driver be doing RAID. Like
>>>> striped or mirrored or RAID5 parity.
>>>>
>>>> This scheme is designed to allow "RAID Migration". Have a single
>>>> drive with C: on it. Add a blank drive to machine. Make an array
>>>> from original C: and blank drive, and it automatically converts
>>>> it to RAID 0/1/5 and so on. This was intended to solve the problem
>>>> which existed previously, where it was so hard to get a RAID
>>>> install going.
>>>>
>>>> If the Intel driver was fouled up, why would the boot even finish ?
>>>> If you only had a single hard drive, the hard drive would be
>>>> running as AHCI (using the AHCI file in the Intel RAID folder).
>>>>
>>>> *******
>>>>
>>>> OMG. The stupid stuff you find on teh Internet.
>>>>
>>>> https://wlmail.wordpress.com/esent-errors/
>>>>
>>>>    0x8E5E0247 SectorSizeNotSupported  The physical sector size
>>>> reported
>>>>                                       by the disk subsystem, is
>>>> unsupported
>>>>                                       by ESE for a specific file type.
>>>>
>>>> That would be stuff like 512e (emulation), 512n (native), 4Kn and
>>>> so on.
>>>>
>>>>    512e  4096 physical  512 virtual    Only an issue with WinXP
>>>> alignment, OK on Win7 alignment
>>>>    512n   512 physical  512 virtual    Bulletproof compatibility,
>>>> traditional size, WDC Gold small disks
>>>>    4Kn   4096 physical 4096 virtual    Do not buy, use, or mix
>>>> these with desktop PCs... Trouble awaits!
>>>>                                        May need to buy
>>>> different/new tools, for stuff.
>>>>
>>>> 512e is a lot more common than 512n today, and while some brands
>>>> may have 512n,
>>>> it could disappear in any model year. One of the drives in the Win7
>>>> machine
>>>> is a 512n, and suited to seamless usage in WinXP (which can no
>>>> longer happen here).
>>>>
>>>> These have been around long enough, this is surely some kind of
>>>> bogus error.
>>>>
>>>> https://mskb.pkisolutions.com/kb/2470478
>>>>
>>>>    "ESENT expects that the reported physical sector size remains
>>>> *consistent*
>>>>     between sessions. When the ESENT database re-initializes, it
>>>> makes sure
>>>>     that the physical sector size that the database was created for
>>>> and the
>>>>     current reported physical sector size are the same. When the
>>>> sector sizes
>>>>     are different, ESENT reports an error."
>>>>
>>>> The ST31000528AS appears to be 512n. The spec is dated 2009, Wiki says
>>>> advanced format drives were around 2010. That's why the spec does not
>>>> contain the nomenclature expected.
>>>>
>>>>    As administrator:
>>>>
>>>>    fsutil fsinfo ntfsinfo c:
>>>>
>>>>    NTFS Volume Serial Number :        0x1234abcd1234abcd
>>>>    NTFS Version      :                3.1
>>>>    Bytes Per Sector  :                512            <===
>>>> LogicalSector  \__ a "512 emulated"
>>>>    Bytes Per Physical Sector :        4096           <===
>>>> PhysicalSector /   drive
>>>>    Bytes Per Cluster :                4096  (4 KB) file system
>>>> storage unit
>>>>    Bytes Per FileRecord Segment    :  1024           <=== $MFT slot
>>>> size
>>>>
>>>> Check and see what info you can find on yours. Yours is currently
>>>> likely to be 512 512 4096 1024. If you were in RAID mode (two drives),
>>>> I don't know what they stuff in those fields for that.
>>>>
>>>> *******
>>>>
>>>> If the config is working in Windows 10, including calendar, could
>>>> you bring the materials back to the Win7 machine ?
>>>>
>>>> Winston seems to know a lot of trivia for WLM, perhaps he will
>>>> have a suggestion.
>>>>
>>>> But on the surface, this is just... ridiculous. Absurd. Like a
>>>> steam powered rocket.
>>>> This can't be it, can it?
>>>>
>>>>    Paul
>>>>
>>> You got me thinking about the drive set up on the Windows 7 machine.
>>> There are two SATA drives attached as RAID:
>>>
>>> 1) 1TB SATA Seagate ST31000528AS, my C:, with OS, programs, data, etc.
>>>
>>> 2) 1TB SATA WDC10EZEX-08WN4A0, my F:, a clone of C: from when C: was
>>> reinstalled fresh with Windows 7 and programs, including WLM, about
>>> three years ago;
>>>
>>> I tried booting from 2), and launching WLM.  No luck, same error.
>>>
>>> I disconnected power from 2) and machine booted from 1).  WLM
>>> launched just fine.  I've shut down WLM many times, and it comes
>>> back up fine. I guess next step is to re-attach 2) and see if the
>>> error returns when WLM is launched from 1).
>>>
>>> If it's 2) that's suddenly causing the error, it may explain why a
>>> restored image of an 'error free' backup STILL gave the error.
>>>
>>>
>> I forgot to say that I've rebooted with only 1) connected, and WLM
>> still comes up fine.  In the morning, I'll re-attach 2), boot from
>> 1), and see if I get the WLM error message.
>
> It would help to run the
>
>    fsutil fsinfo ntfsinfo c:
>
> as administrator, to see if there is any pattern to the
>
>    Bytes Per Sector:            512
>    Bytes Per Physical Sector:  4096
>
> thing or not. I find it hard to believe, that a RAID driver
> would be substituting the stripe size for the Physical Sector
> value. The stripe size could be 64KB or 128KB or so.
>
> You would want to see what values make it work, and what
> values make it crash with that particular error code.
>
>    Paul
>
Like clockwork, if only C: is attached, WLM works without error. If I
also attach G: (the WD drive), WLM fails everytime.  If I detach G:, WLM
works without error.

On 7/20/2022 6:47 PM, Boris wrote:

> Like clockwork, if only C: is attached, WLM works without error. If I also attach G: (the WD drive), WLM fails everytime.  If I detach G:, WLM works without error.
>
> The results of fsutil...
>
> with only C: attached
>
> https://postimg.cc/9wYdbmF0
>
>
> with C: and G: attached
>
> https://postimg.cc/WFsxS3sq

And one of the fields says <not supported> :-) Ha!

So I guess there's no passthru to get the value
via that driver or something.

Paul

On 7/20/2022 3:55 PM, Paul wrote:
> On 7/20/2022 6:47 PM, Boris wrote:
>
>> Like clockwork, if only C: is attached, WLM works without error. If I
>> also attach G: (the WD drive), WLM fails everytime.  If I detach G:,
>> WLM works without error.
>>
>> The results of fsutil...
>>
>> with only C: attached
>>
>> https://postimg.cc/9wYdbmF0
>>
>>
>> with C: and G: attached
>>
>> https://postimg.cc/WFsxS3sq
>
> And one of the fields says <not supported> :-) Ha!
>
> So I guess there's no passthru to get the value
> via that driver or something.
>
>    Paul
>
>
So I guess there's no passthru to get the value
via that driver or something.

Paul

It's odd that "Not Supported" appears for C: only if G: is also
attached.  If G: is not attached, the Bytes Per Physical Sector is 512.
Another mystery (to me) is if I attach only G:,Bytes Per Physical Sector
now shows "Not Supported", shown here.

https://postimg.cc/sGr4dnnM

The prompt shows C: but it is the G: (WD with Windows 7 OS).  I also
noticed that the RM identifier has changed to the same as what is shown
for the C: when both C: and G: are attached.  The RM Identifier seems to
follow with the "Not Supported" paramater.  Do you know what this is about?

On 7/20/2022 8:09 PM, Boris wrote:
> On 7/20/2022 3:55 PM, Paul wrote:
>> On 7/20/2022 6:47 PM, Boris wrote:
>>
>>> Like clockwork, if only C: is attached, WLM works without error. If I also attach G: (the WD drive), WLM fails everytime.  If I detach G:, WLM works without error.
>>>
>>> The results of fsutil...
>>>
>>> with only C: attached
>>>
>>> https://postimg.cc/9wYdbmF0
>>>
>>>
>>> with C: and G: attached
>>>
>>> https://postimg.cc/WFsxS3sq
>>
>> And one of the fields says <not supported> :-) Ha!
>>
>> So I guess there's no passthru to get the value
>> via that driver or something.
>>
>>    Paul
>>
>>
> So I guess there's no passthru to get the value
> via that driver or something.
>
> Paul
>
>
> It's odd that "Not Supported" appears for C: only if G: is also attached.  If G: is not attached, the Bytes Per Physical Sector is 512.
> Another mystery (to me) is if I attach only G:,Bytes Per Physical Sector now shows "Not Supported", shown here.
>
> https://postimg.cc/sGr4dnnM
>
> The prompt shows C: but it is the G: (WD with Windows 7 OS).  I also noticed that the RM identifier has changed to the same as what is shown for the C: when both C: and G: are attached.  The RM Identifier seems to follow with the "Not Supported" paramater.  Do you know what this is about?
>

RM Identifier = Resource Management Identifier and this appears
to be some deprecated thing once used by Windows Server for some reason.

Do you know if your system is actually trying to run a RAID array right now ?
It could be, that some of the side effects here, are from an array
changing states from "working" to "degraded" or "failed" kind of thing.
If it was "failed", of course the thing would not be accessible
from a data perspective. The way things are going for you, suggests
you may have been playing with RAID1 Mirror at some time.

I'm just trying to understand what is going on here :-)

If the disks were JBOD AHCI, the disks should respond
like ordinary disks, except the total volume size might
become smaller by one cylinder or so. As the metadata for
the Intel driver might be at the very end of the disk drive.
Each disk drive in a pair would have metadata declaring
which partner it was. For example, on striped RAID0, one
drive is odd, one drive is even, counting from zero. The
"stripes" laid down, alternate between drives, but you
need to know which drive is the zero drive. Using metadata,
if the user unplugs the drives and puts them on different
SATA ports, the driver can still figure out which drive is
which.

If I was running RAID on the Intel machine right now, I
could move the disks to the AMD machine and clean off the
metadata with a "diskpart" "clean". If I had formed a RAID
array on the AMD machine, I could move the array to the
Intel machine and clean off the AMD metadata (the AMD metadata
being in Promise Technology format under license).

On my WinXP machine with the Core2 processor, I could clean
off Intel metadata using the JMicron port located on
the motherboard of the same machine. Generally, with brand
changes, you get the opportunity to "clean" a disk properly.
Only AMD and Promise Technology are interchangeable, because
AMD bought software from Promise for usage as AMD RAID.

And that's only if you're having trouble "shaking" some RAID
habit :-) To clean RAID drives properly is difficult but not impossible.

Using Macrium, you may be able to back up C: off one kind
of volume, and restore it to some other kind of volume,
but with our newly acquired knowledge about ESE Jet database,
with some risk that our WLM email/calendar won't work.

When we have a configuration that works, using the fsutil
command to note the logical sector size and physical sector
size, gives us a hint as to what alternate disk driving format
is going to work.

Normally, before learning of your problem, I would "assume" that
software does not care about the details of physical storage.

But now that this ESE Jet Database issue has been highlighted,
I can see all sorts of opportunities for "misfortune".

This is not the first time an ESE issue has come up,
where I was wishing for some utilities to "clean" the database.
But if we have no tools, nothing to "reset" the physical sector
value and the "preferred schema" the database uses for storage,
why that's like trying to work with handcuffs on. I think
something on Linux can read a database, after a fashion,
but I don't know if there's such a thing as a complete
maintenance suite there. I doubt it. Normally reading
stuff is "a bare minimum of support".

There is supposed to be a maintenance tool on Windows Server,
but I don't know how we'd get our hands on that.

*******

With RAID disks in place:

1) The BIOS can have a RAID Management screen, for examining
what metadata says about any arrays that have been created.
The BIOS grays out the screen, for any aspects that are not
currently valid. The interface never says "bad Monkey" -- it's
just certain items in the screen are grayed out so you cannot
use them, and you use the "gray color" as a hint as to what
the problem might be. Like if I notice "the option to make a RAID0
right now is grayed out", I have to hit the user manual and see
what reasons exist for denial of RAID0 mode.

2) Installing a RAID driver is one thing. The Windows "Intel RAID Management Package"
provides the same kinds of details as (1), except in the comfort
of the Windows desktop. You can rebuild degraded arrays in there,
note that an array is in a degraded state, and so on. Anyone
doing RAID configs, should have the RAID Management Package installed.
The interface is more likely to pop up a balloon "you cannot make
a RAID0 right now, because the allowed number of RAID0 arrays has
been exceeded". There is more room for "good Monkey", or "bad Monkey"
messages in the Windows one.

But, if you're making your very first RAID0, followed by popping
in a Windows 7 installer disc and doing a clean OS install, the BIOS screen
is the only viable option. With brand new disks with no metadata on
them, the screen should come up immediately allowing a RAID0 to be
formed. Disks that have "histories", that BIOS screen could be
twisted six ways from Sunday, and nothing but misery results for
the computer operator. Thus, one of the first skills it pays to
learn, is how to clean a disk properly so all metadata is removed.

Switching SATA port brands, helps. The Intel driver, does not parse,
honor or respect, the Promise metadata structure. And vice versa.
That's the beauty of changing SATA port brands. Instant amnesia.
And then the clean process will work.

(admin terminal or admin command prompt)

diskpart
list disk
select disk 1
clean all # if Intel RAID disk on AMD SATA port, cleans last cylinder
# which would otherwise be "out of reach".
# A full clean could take 1-6 hours for small disks.
exit

But you do not resort to cleaning, until you have your Macrium backup
stored on a USB disk somewhere.

Summary: RAID == "FCUK, Now I Have A Hobby" :-)

And it's not even as much fun as stamp collecting.

I DO NOT recommend RAID for the users of this group.
You don't need that hobby. I've done RAIDs here, to learn
how to use them, but in no way am I "attracted" to RAID
enough to seek out RAID every time a new machine shows up.
I only did enough RAID so I could try to help. Like, now,
I know how to clean a disk.

HTH,
Paul

Paul <nospam@needed.invalid> wrote in news:tbaahn$nb2$1@gioia.aioe.org:

> On 7/20/2022 8:09 PM, Boris wrote:
>> On 7/20/2022 3:55 PM, Paul wrote:
>>> On 7/20/2022 6:47 PM, Boris wrote:
>>>
>>>> Like clockwork, if only C: is attached, WLM works without error. If
>>>> I also attach G: (the WD drive), WLM fails everytime.  If I detach
>>>> G:, WLM works without error.
>>>>
>>>> The results of fsutil...
>>>>
>>>> with only C: attached
>>>>
>>>> https://postimg.cc/9wYdbmF0
>>>>
>>>>
>>>> with C: and G: attached
>>>>
>>>> https://postimg.cc/WFsxS3sq
>>>
>>> And one of the fields says <not supported> :-) Ha!
>>>
>>> So I guess there's no passthru to get the value
>>> via that driver or something.
>>>
>>>    Paul
>>>
>>>
>> So I guess there's no passthru to get the value
>> via that driver or something.
>>
>> Paul
>>
>>
>> It's odd that "Not Supported" appears for C: only if G: is also
>> attached.  If G: is not attached, the Bytes Per Physical Sector is
>> 512. Another mystery (to me) is if I attach only G:,Bytes Per
>> Physical Sector now shows "Not Supported", shown here.
>>
>> https://postimg.cc/sGr4dnnM
>>
>> The prompt shows C: but it is the G: (WD with Windows 7 OS).  I also
>> noticed that the RM identifier has changed to the same as what is
>> shown for the C: when both C: and G: are attached.  The RM
>> Identifier seems to follow with the "Not Supported" paramater.  Do
>> you know what this is about?
>>
>
> RM Identifier = Resource Management Identifier and this appears
> to be some deprecated thing once used by Windows Server for some
> reason.
>
> Do you know if your system is actually trying to run a RAID array
> right now ? It could be, that some of the side effects here, are from
> an array changing states from "working" to "degraded" or "failed" kind
> of thing. If it was "failed", of course the thing would not be
> accessible from a data perspective. The way things are going for you,
> suggests you may have been playing with RAID1 Mirror at some time.
>
> I'm just trying to understand what is going on here :-)

I'm away from the machine now, and won't be back to it for many days, so
I can't send any screenshots, but I can say that the the thing shipped
with one disk in 2010, and the POST screen showed it was set up as RAID.
Not RAID0 or RAID1, just RAID. I added a second disk some years later,
and POST showed RAID for both. After reading your note here, I decided
to see if I could change the RAID setting. In BIOS, my choice was either
RAID or ATA. I changed both to ATA, and rebooted. The machine would not
boot up. I entered BIOS again, and set back to RAID. The machine still
would not boot up, and Windows presented me with the System Recovery
Options screen.

https://postimg.cc/jL4ShHGM

I choose Startup Repair, but after a long while, the system said it could
not repair itself. Next I choose System Restore. All of my restore
points were gone, and the date of the point to be used made me think this
isn't a good point. But, it did work. Both POST and BIOS showed RAID.

I will leave it as is, and just not connect the WD drive causing the WLM
error.

>
> If the disks were JBOD AHCI, the disks should respond
> like ordinary disks, except the total volume size might
> become smaller by one cylinder or so. As the metadata for
> the Intel driver might be at the very end of the disk drive.
> Each disk drive in a pair would have metadata declaring
> which partner it was. For example, on striped RAID0, one
> drive is odd, one drive is even, counting from zero. The
> "stripes" laid down, alternate between drives, but you
> need to know which drive is the zero drive. Using metadata,
> if the user unplugs the drives and puts them on different
> SATA ports, the driver can still figure out which drive is
> which.
>
> If I was running RAID on the Intel machine right now, I
> could move the disks to the AMD machine and clean off the
> metadata with a "diskpart" "clean". If I had formed a RAID
> array on the AMD machine, I could move the array to the
> Intel machine and clean off the AMD metadata (the AMD metadata
> being in Promise Technology format under license).
>
> On my WinXP machine with the Core2 processor, I could clean
> off Intel metadata using the JMicron port located on
> the motherboard of the same machine. Generally, with brand
> changes, you get the opportunity to "clean" a disk properly.
> Only AMD and Promise Technology are interchangeable, because
> AMD bought software from Promise for usage as AMD RAID.
>
> And that's only if you're having trouble "shaking" some RAID
> habit :-) To clean RAID drives properly is difficult but not
> impossible.
>
> Using Macrium, you may be able to back up C: off one kind
> of volume, and restore it to some other kind of volume,
> but with our newly acquired knowledge about ESE Jet database,
> with some risk that our WLM email/calendar won't work.
>
> When we have a configuration that works, using the fsutil
> command to note the logical sector size and physical sector
> size, gives us a hint as to what alternate disk driving format
> is going to work.
>
> Normally, before learning of your problem, I would "assume" that
> software does not care about the details of physical storage.
>
> But now that this ESE Jet Database issue has been highlighted,
> I can see all sorts of opportunities for "misfortune".
>
> This is not the first time an ESE issue has come up,
> where I was wishing for some utilities to "clean" the database.
> But if we have no tools, nothing to "reset" the physical sector
> value and the "preferred schema" the database uses for storage,
> why that's like trying to work with handcuffs on. I think
> something on Linux can read a database, after a fashion,
> but I don't know if there's such a thing as a complete
> maintenance suite there. I doubt it. Normally reading
> stuff is "a bare minimum of support".
>
> There is supposed to be a maintenance tool on Windows Server,
> but I don't know how we'd get our hands on that.
>
> *******
>
> With RAID disks in place:
>
> 1) The BIOS can have a RAID Management screen, for examining
> what metadata says about any arrays that have been created.
> The BIOS grays out the screen, for any aspects that are not
> currently valid. The interface never says "bad Monkey" -- it's
> just certain items in the screen are grayed out so you cannot
> use them, and you use the "gray color" as a hint as to what
> the problem might be. Like if I notice "the option to make a RAID0
> right now is grayed out", I have to hit the user manual and see
> what reasons exist for denial of RAID0 mode.
>
> 2) Installing a RAID driver is one thing. The Windows "Intel RAID
> Management Package"
> provides the same kinds of details as (1), except in the comfort
> of the Windows desktop. You can rebuild degraded arrays in there,
> note that an array is in a degraded state, and so on. Anyone
> doing RAID configs, should have the RAID Management Package
> installed. The interface is more likely to pop up a balloon "you
> cannot make a RAID0 right now, because the allowed number of RAID0
> arrays has been exceeded". There is more room for "good Monkey",
> or "bad Monkey" messages in the Windows one.
>
> But, if you're making your very first RAID0, followed by popping
> in a Windows 7 installer disc and doing a clean OS install, the BIOS
> screen is the only viable option. With brand new disks with no
> metadata on them, the screen should come up immediately allowing a
> RAID0 to be formed. Disks that have "histories", that BIOS screen
> could be twisted six ways from Sunday, and nothing but misery results
> for the computer operator. Thus, one of the first skills it pays to
> learn, is how to clean a disk properly so all metadata is removed.
>
> Switching SATA port brands, helps. The Intel driver, does not parse,
> honor or respect, the Promise metadata structure. And vice versa.
> That's the beauty of changing SATA port brands. Instant amnesia.
> And then the clean process will work.
>
> (admin terminal or admin command prompt)
>
> diskpart
> list disk
> select disk 1
> clean all # if Intel RAID disk on AMD SATA port, cleans last
> cylinder
> # which would otherwise be "out of reach".
> # A full clean could take 1-6 hours for small
> disks.
> exit
>
> But you do not resort to cleaning, until you have your Macrium backup
> stored on a USB disk somewhere.
>
> Summary: RAID == "FCUK, Now I Have A Hobby" :-)
>
> And it's not even as much fun as stamp collecting.
>
> I DO NOT recommend RAID for the users of this group.
> You don't need that hobby. I've done RAIDs here, to learn
> how to use them, but in no way am I "attracted" to RAID
> enough to seek out RAID every time a new machine shows up.
> I only did enough RAID so I could try to help. Like, now,
> I know how to clean a disk.
>
> HTH,
> Paul

On 7/22/2022 8:00 PM, Boris wrote:

> I'm away from the machine now, and won't be back to it for many days, so
> I can't send any screenshots, but I can say that the the thing shipped
> with one disk in 2010, and the POST screen showed it was set up as RAID.
> Not RAID0 or RAID1, just RAID. I added a second disk some years later,
> and POST showed RAID for both. After reading your note here, I decided
> to see if I could change the RAID setting. In BIOS, my choice was either
> RAID or ATA. I changed both to ATA, and rebooted. The machine would not
> boot up. I entered BIOS again, and set back to RAID. The machine still
> would not boot up, and Windows presented me with the System Recovery
> Options screen.
>
> https://postimg.cc/jL4ShHGM
>
> I choose Startup Repair, but after a long while, the system said it could
> not repair itself. Next I choose System Restore. All of my restore
> points were gone, and the date of the point to be used made me think this
> isn't a good point. But, it did work. Both POST and BIOS showed RAID.
>
> I will leave it as is, and just not connect the WD drive causing the WLM
> error.
>

This sounds a lot like my "RAID Ready" Dell.

See if you can find the BIOS RAID screen, and see
what it has "assumed" about your drives.

It should *not* join the drives, without user input.
But maybe that's what it has done.

You see, on my Dell, it has a lot of trouble with
boot selection, and this has a lot to do with its
RAID fetish. You have to "turn ports ON and OFF" to
get anything done on the box. None of my home builds
ever require doing things like that.

*******

I looked for more info on the ESE database file format,
and the only parameter in there is "Pagesize", which
is 2K, 16K, 32K. None of these values, particular smack
of "storage subsystem" values. It's probably the
<Not Supported> coming back from a request, which is
upsetting it. As I could not find any mention of it
memorizing logical and physical sector size as such.
The web page I was using, is someone who was reverse
engineering the format.

Paul

Paul <nospam@needed.invalid> wrote in news:tbfg09$dnn$1@gioia.aioe.org:

> On 7/22/2022 8:00 PM, Boris wrote:
>
>> I'm away from the machine now, and won't be back to it for many days,
>> so I can't send any screenshots, but I can say that the the thing
>> shipped with one disk in 2010, and the POST screen showed it was set
>> up as RAID. Not RAID0 or RAID1, just RAID. I added a second disk
>> some years later, and POST showed RAID for both. After reading your
>> note here, I decided to see if I could change the RAID setting. In
>> BIOS, my choice was either RAID or ATA. I changed both to ATA, and
>> rebooted. The machine would not boot up. I entered BIOS again, and
>> set back to RAID. The machine still would not boot up, and Windows
>> presented me with the System Recovery Options screen.
>>
>> https://postimg.cc/jL4ShHGM
>>
>> I choose Startup Repair, but after a long while, the system said it
>> could not repair itself. Next I choose System Restore. All of my
>> restore points were gone, and the date of the point to be used made
>> me think this isn't a good point. But, it did work. Both POST and
>> BIOS showed RAID.
>>
>> I will leave it as is, and just not connect the WD drive causing the
>> WLM error.
>>
>
> This sounds a lot like my "RAID Ready" Dell.

Funny you should mention that. This machine is a Dell XPS 8100. It
shipped running in RAID, back in 2010, with a single drive. I still have
two other working Dells around the house, rarely used, both running XP
SP3. I'm going to check to see how those drives are set up. But, can't
do until I'm home after the 29th.
>
> See if you can find the BIOS RAID screen, and see
> what it has "assumed" about your drives.
>
> It should *not* join the drives, without user input.
> But maybe that's what it has done.
>
> You see, on my Dell, it has a lot of trouble with
> boot selection, and this has a lot to do with its
> RAID fetish. You have to "turn ports ON and OFF" to
> get anything done on the box. None of my home builds
> ever require doing things like that.
>
> *******
>
> I looked for more info on the ESE database file format,
> and the only parameter in there is "Pagesize", which
> is 2K, 16K, 32K. None of these values, particular smack
> of "storage subsystem" values. It's probably the
> <Not Supported> coming back from a request, which is
> upsetting it. As I could not find any mention of it
> memorizing logical and physical sector size as such.
> The web page I was using, is someone who was reverse
> engineering the format.
>
> Paul
>