Image for Windows 3.07

User discussion and information resource forum for Image products.
DrTeeth
Posts: 1289
Joined: Fri Aug 12, 2011 6:58 pm

Image for Windows 3.07

Post by DrTeeth »

Hi,

First, thanks for increasing the number of PCs on which we can use
your programs.

I notice from the change log that there have been quite a few changes
to the metadata system. Do any of the changes address the issue that I
had, viz when using metadata it took sometimes well over 30 mins for
the backup to start rather than the less than 30 seconds without it?

Because of the behaviour of v2, I got into the habit of waiting until
the backup had started before leaving my PC. I run the backup at night
when I have finished playing (ham radio at the moment) before
retiring.
--
Cheers,

DrT

** Stress - the condition brought about by having to
** resist the temptation to beat the living daylights
** out of someone who richly deserves it.
Logman
Posts: 112
Joined: Fri Jan 16, 2015 6:10 pm

Re: Image for Windows 3.07

Post by Logman »

DrTeeth wrote:
> Hi,
>
> First, thanks for increasing the number of PCs on which we can use
> your programs.
>

That is great isn't it ?
Much appreciated 8-)
TeraByte Support
Posts: 3598
Joined: Thu May 05, 2011 10:37 pm

Re: Image for Windows 3.07

Post by TeraByte Support »

probably not, not sure how it could take that long unless drive very slow,
those may benefit from device caching but still not on since made no
difference on testing systems, those systems only took a minute or so for
even very large number of files (a few minutes if enable to include dir
contents).

In general you shouldn't use the metadata option unless you understand
it's limitations or impact. Also if you do use it and do live backups ensure
you use VSS with it.

"DrTeeth" wrote in message news:13836@public.image...

Hi,

First, thanks for increasing the number of PCs on which we can use
your programs.

I notice from the change log that there have been quite a few changes
to the metadata system. Do any of the changes address the issue that I
had, viz when using metadata it took sometimes well over 30 mins for
the backup to start rather than the less than 30 seconds without it?
DrTeeth
Posts: 1289
Joined: Fri Aug 12, 2011 6:58 pm

Re: Image for Windows 3.07

Post by DrTeeth »

On Sat, 27 May 2017 08:30:21 PDT, just as I was about to take a herb,
TeraByte Support disturbed my reverie and wrote:

>In general you shouldn't use the metadata option unless you understand
>it's limitations or impact. Also if you do use it and do live backups ensure
>you use VSS with it.

Hi,

The drives are not slow, they are SATA 3 drives, but connected to a
SATA 2 interface. No other issues here that could be related to a
drive's slowness. I am happy to help you investigate/test my system
for your information. Please, if interested, send me an email if you
can find my address on your system. If not please let me know and I
will email you.

I wonder if you could explain why VSS should be used instead of your
excellent PHYLock? What are the limitations and impact? I was under
the impression that it would speed up differential/incremental backups
- no more, no less. I am asking for information and not to be "funny",
though I may come across that way.
--
Cheers,

DrT

"If you want to find out what is wrong
with democracy, spend five minutes with
the average voter." - Winston Churchill
Muad'Dib
Posts: 110
Joined: Mon Nov 21, 2011 12:23 pm

Re: Image for Windows 3.07

Post by Muad'Dib »

DrTeeth wrote: "First, thanks for increasing the number of PCs on which we can use your programs."

I agree! Now I can let my pets use IFW for their computers (I'm a real person, they are real pets who are members of my immediate family, and they live with me, so it's all legit!). :mrgreen:

That said, a couple of website corrections need to be made:

1) While the downloaded license now shows you can use it for up to 5 computers for personal/home use, all the website purchase pages (for example http://www.terabyteunlimited.com/purcha ... indows.htm) still say each personal license is limited to 3 computers.

2) There's a version error on the download page for IFW: http://www.terabyteunlimited.com/downlo ... indows.htm still shows the version as 3.06a, though May 26th is correctly shown as the latest availability date (IFD, IFL and BIBM pages have the correct version # and date posted).
TeraByte Support
Posts: 3598
Joined: Thu May 05, 2011 10:37 pm

Re: Image for Windows 3.07

Post by TeraByte Support »

ok, thanks.

"Muad'Dib" wrote in message news:13843@public.image...

DrTeeth wrote: "First, thanks for increasing the number of PCs on which we
can use your programs."

I agree! Now I can let my pets use IFW for their computers (I'm a real
person, they are real pets who are members of my immediate family, and they
live with me, so it's all legit!).

![:mrgreen:]({SMILIES_PATH}/icon_mrgreen.gif)

That said, a couple of website corrections need to be made:

1) While the downloaded license now shows you can use it for up to 5
computers for personal/home use, all the website purchase pages (for example

[
http://www.terabyteunlimited.com/purcha ... indows.htm
](http://www.terabyteunlimited.com/purchase-image-for-windows.htm)

) still say each personal license is limited to 3 computers.

2) There's a version error on the download page for IFW:

[
http://www.terabyteunlimited.com/downlo ... indows.htm
](http://www.terabyteunlimited.com/downloads-image-for-windows.htm)

still shows the version as 3.06a, though May 26th is correctly shown as the
latest availability date (IFD, IFL and BIBM pages have the correct version #
and date posted).

TeraByte Support
Posts: 3598
Joined: Thu May 05, 2011 10:37 pm

Re: Image for Windows 3.07

Post by TeraByte Support »

>I wonder if you could explain why VSS should be used instead of your
>excellent PHYLock? What are the limitations and impact?

For normal backups that don't use /usemd, you can continue to use either.
VSS is recommended for /usemd because of open files that stay open nearly
all the time that have only updates over existing data (no
additions/truncations) and which metadata is not updated until the file is
closed. In windows 10, there are a lot of these type of caching files which
are also set to not update any timestamp information even when the handle is
closed. For those files, the only metadata change may be the USN when the
USN Journal is enabled on the volume. Sticking with those type of files
(ones where there are no timestamp updates at all, and only a USN update
when the USN Journal is enabled); when VSS starts, it triggers the USN to
update on the open file. When PHYLock starts and even though requests both
locking and volume flush on the volume, Windows ignores and does not update
the USN on the files. The difference can be explained with an example:
Take file X that has no timestamp updates, file updates with no appending or
truncation, does have USN updates. Backup A includes file X, Backup B, C
the file X was not changed and therefore data used from A. Backup D, File
X is open and has been updated but the USN is still as it was in A, B, C.
The file data is backed up both using VSS and PHYLock, the difference is in
the VSS version USN is different than ABC, PHYlock the same as ABC. Now
file X is closed with no other updates other than what was already done
prior. Backup E VSS version would skip the data and use what is in D,
PHYLock version would see change and backup the data in to E. Now it's
time to restore, if we just booted and restored back to B, all fine in all
versions, however, if you first restored D version then right after that the
B version, the D restored with VSS version will have a USN different than B
so will restore file X data from B, PHYLock version has the same USN so
ignores restore file X data. So the PHYLock restored version of B has D's
data for file X. The byte-for-byte validation would confirm and fail since
it's done over the entire data. Of course these are special files, special
restores, but to ensure best metadata state, VSS should be used (If Windows
were changed to flush the metadata of open handles on lock or volume flush
request, then the difference goes away at the metadata level and either
would give the same metadata results).


Brian K
Posts: 2214
Joined: Fri Aug 12, 2011 1:11 am
Location: NSW, Australia

Re: Image for Windows 3.07

Post by Brian K »

I'm trying to get this clear in my head. Let's say you did daily backups A, B, C, D and E. Each backup was done with VSS and then PHYLock. A few weeks later you decide to restore (metadata option) the B backups. Would the restored result be the same if you used the VSS image or the PHYLock image. I'm ignoring the files that might not be backed up by a VSS image.
TeraByte Support
Posts: 3598
Joined: Thu May 05, 2011 10:37 pm

Re: Image for Windows 3.07

Post by TeraByte Support »

yes, because you have to reboot to restore (or lock the drive) which will
close the file and update the USN metadata (most files also timestamp, but
example given to files that don't update timestamp) so there will be a
difference in the metadata on restore (if it was actually changed) so it
would put that data back. However if there was an abnormal bsod or crash
or power off, and metadata never updated, then it be in the same condition
as the other example given , but in that case of a system crash or
corruption you should normally do full restore anyway not using /usemd
(unless you know it was something you just installed).

but then again, if you restored b which is fine, if you then later restore say d,
if that file is not updated since being restored, then restoring d won't change
that file (reason same as back to the prior example).

"Brian K" wrote in message news:13851@public.image...

I'm trying to get this clear in my head. Let's say you did daily backups A,
B, C, D and E. Each backup was done with VSS and then PHYLock. A few weeks
later you decide to restore (metadata option) the B backups. Would the
restored result be the same if you used the VSS image or the PHYLock image.
I'm ignoring the files that might not be backed up by a VSS image.
Walter B
Posts: 93
Joined: Fri Aug 12, 2011 1:48 pm

Re: Image for Windows 3.07

Post by Walter B »

Does this all boil down to: if you are using Win 10 Pro X64, then use
VSS?


On Mon, 29 May 2017 13:41:04 PDT, "TeraByte Support"
wrote:

>>I wonder if you could explain why VSS should be used instead of your
>>excellent PHYLock? What are the limitations and impact?
>
>For normal backups that don't use /usemd, you can continue to use either.
>VSS is recommended for /usemd because of open files that stay open nearly
>all the time that have only updates over existing data (no
>additions/truncations) and which metadata is not updated until the file is
>closed. In windows 10, there are a lot of these type of caching files which
>are also set to not update any timestamp information even when the handle is
>closed. For those files, the only metadata change may be the USN when the
>USN Journal is enabled on the volume. Sticking with those type of files
>(ones where there are no timestamp updates at all, and only a USN update
>when the USN Journal is enabled); when VSS starts, it triggers the USN to
>update on the open file. When PHYLock starts and even though requests both
>locking and volume flush on the volume, Windows ignores and does not update
>the USN on the files. The difference can be explained with an example:
>Take file X that has no timestamp updates, file updates with no appending or
>truncation, does have USN updates. Backup A includes file X, Backup B, C
>the file X was not changed and therefore data used from A. Backup D, File
>X is open and has been updated but the USN is still as it was in A, B, C.
>The file data is backed up both using VSS and PHYLock, the difference is in
>the VSS version USN is different than ABC, PHYlock the same as ABC. Now
>file X is closed with no other updates other than what was already done
>prior. Backup E VSS version would skip the data and use what is in D,
>PHYLock version would see change and backup the data in to E. Now it's
>time to restore, if we just booted and restored back to B, all fine in all
>versions, however, if you first restored D version then right after that the
>B version, the D restored with VSS version will have a USN different than B
>so will restore file X data from B, PHYLock version has the same USN so
>ignores restore file X data. So the PHYLock restored version of B has D's
>data for file X. The byte-for-byte validation would confirm and fail since
>it's done over the entire data. Of course these are special files, special
>restores, but to ensure best metadata state, VSS should be used (If Windows
>were changed to flush the metadata of open handles on lock or volume flush
>request, then the difference goes away at the metadata level and either
>would give the same metadata results).
>

Walter B
Post Reply