A step-by-step example of how to use this
function can be found here.
Why clone a disk?
The first thing a professional data recovery engineer will attempt before
performing any repair or recovery operation, is cloning the problem disk. No
matter how simple the problem may seem at first sight, a cloned disk offers a
good safety net in case the recovery doesn't go according to plan.
Cloning a disk first offers several key advantages:
you can perform your repairs on
the clone. If the repairs don't work out as planned, you still have the original
disk in its unmodified state, so you could create another clone to try again.
in case of bad sectors, repairs can not
be made on the original (bad) disk. If bad sectors exist in areas on the disk
that contain disk structures, repairing these structures is going to be
impossible because you can't write to bad sectors. Cloning the bad disk to
a good disk will give you a much better chance of repairing those damaged
Be prepared for the fact that cloning a bad disk to a good disk can take a
considerable amount of time.
Having said this, you can safely make repairs on a physically healthy disk using
repairs do not bring the desired results (access to your data) or even make
matters worse, you can undo the repairs you have made.
If you can be certain that the disk needing repairs does not have any physical
defects (bad sectors) and you don't have a spare disk handy for cloning, you can
forego the option to clone a disk. You can check a disk for physical defects by
running a surface scan, or by running a less
time-consuming health check.
If you're performing a
partition table repair or a boot sector rebuild and the disk is in good
(physical) condition, a clone is not needed.
If the disk shows signs of having bad sectors, a clone should be made as
soon as possible: not only to facilitate repairs (if needed) but also to copy
the data from the bad disk before the disk fails altogether.
Many of the 'conventional' cloning products, disk copiers and disk imagers are
not particularly good at handling disks containing bad sectors or corrupt disk structures.
The DiskPatch cloning feature is designed to handle disks that contain unreadable areas, and to ignore any damage
in disk structures such as partition tables and file systems.
Keep in mind
though that the DiskPatch clone feature does not offer the functions that
'conventional' imaging tools offer, like resizing partitions; clone is meant to
Important: If a disk is cloned
for forensic purposes (evidence acquisition) refer to the "Forensics"
page for background information and procedures.
How DiskPatch cloning works
DiskPatch uses an intelligent method to make sure that as much data as possible
gets copied. This is achieved by running the clone using a 2 step procedure:
|Step 1: copy as much data as possible in the quickest possible way. Read retries are ignored to make sure that the first step completes as fast as possible. When a bad sector is encountered the clone process will skip a number of sectors and add the location where the skip starts to a list. This list is used in step 2.
Step 2: clone the areas that were skipped in step 1. Read retries are now used and the sectors are handled in such a way that as much as possible gets copied. Step 2 can be time consuming; depending on settings bad sectors can slow down the clone process severely.
- skipping sectors (step 1) is used
because bad sectors are usually found in groups; skipping more than 1 sector
after running into bad sectors will speed up the clone process. The number
of sectors that should be skipped can be configured in the options screen.
- both steps can be run automatically
(meaning you can run a 'full' clone; step 2 will start automatically when
step 1 has finished), or you can run only step 1 or step 2. Step 2 can only
be run if step 1 has been completed first.
- step 2 only needs to run if sectors
were skipped during step 1. If step 2 starts and no additional sectors need
to be copied, DiskPatch will notify you of this.
- if you restart step 1 after you have aborted step 1 you will have to
start from the beginning again.
- if you abort step 2 you DO NOT
have to run step 1 again. DiskPatch keeps a list of sectors that need to be
copied for step 2 and that list is updated as step 2 runs to completion. If
you abort step 2 before the entire list has been processed, restarting step
2 will simply continue processing that list.
- the bad sector list that is created
during step 1 and used during step 2 is saved on the target disk so exiting
and restarting DiskPatch will not have negative consequences. When the clone
has been completed the list will no longer be on the target disk (it is used
and removed on the fly during step 2 of the clone process) leaving the
cloned disk(area) as an exact copy of the source disk(area).
The advantages of the 2 step clone
process explained through examples:
Consider cloning a disk that is dying rapidly (the number of bad sectors is
growing as the disk is used). If you would clone the disk with a clone tool that
would simply grind its way through the disk from start to finish, it's at all
possible that the actual clone process could cause the disk to seize as it
encounters clusters of bad sectors; you could end up with a partial clone.
If you would clone using the DiskPatch '2 step' clone it could go something like
Step 1 will copy everything that can be read as quickly as possible,
circumventing clusters of bad areas. Step 2 will 'work' the areas that step 1
has skipped and extract as much data as possible. As step 2 does not interfere
with the work that step 1 has already done, the data that has already been
copied during step 1 is safe: if the source disk dies during step 2 then at
least step 1 has copied the bulk of the data. A possible downside to this
scenario is that step 2 could take quite some time. Keep in mind that step 2 can
be aborted and restarted as needed; you could analyze the clone before letting
step 2 run to completion to see if the data you need has already been copied.
You could also change the settings that affect step 2.
Never change the contents of the target disk when you're examining the disk
in between cloning steps.
Now consider cloning a disk that has 1 or 2 bad sectors (in important places,
like the MBR). There's no need to 'grind' through the disk extracting as much as
possible; most sectors can be read normally. Step 1 will complete quickly and
the few bad sectors that will be processed during step 2 will not take very
long. The clone will be completed quickly.
|What happens when a bad sector is encountered.
If DiskPatch encounters a bad sector during a clone operation and the
sector's contents can not be read at all, DiskPatch replaces the sector
data with a text string: CLNSRCRDSERR. Once the clone has finished this can help to locate files that were
affected by the read errors: on the target disk, search through the files and use
CLNSRCRDSERR as a search string.
General clone considerations
if you have a disk that's dying (bad
sectors are appearing) you should keep in mind that each time you read that
bad disk, it could be the last time. Clone this disk as soon as you
you should decide how to handle bad
sectors and read errors during step 2 before you start cloning the disk.
If you use
default settings, each bad sector that is encountered will be read a maximum
of 32 times (read retries) to see if data can be read. This can make step 2 very slow. You
need to decide for yourself where the focus must be:
clone faster and maybe lose some sectors (set read retries lower),
clone slower and get as much as possible (set read retries higher).
You can change the number of read retries in the options screen. Setting this to 0 (zero) will disable
read retries. Setting read retries higher than 64 is generally not a good
Another setting that affects the clone process is the maximum number of read
errors that are allowed before DiskPatch aborts the process (read error
threshold). This affects both steps of the clone. The default
setting for maximum read errors is 32, meaning that after 32 read errors
DiskPatch will interrupt the current process and ask you how to proceed:
ignore further errors, reset the number of errors or disable error checking.
You can change the read error threshold in the options screen. Setting this to 0 (zero) will disable read error checking.
by default step 2 of the clone is done in
reverse, meaning DiskPatch starts at the end of the area that needs to be
cloned and works its way forward. By accessing problematic areas this way
the chance of reading a bad spot is increased, but the procedure may take a
bit more time. If you wish to change this behavior you can change the
corresponding setting in the options screen.
the target disk for the clone should be
at least the same size as the source disk. The target disk is allowed to be
smaller, but DiskPatch will warn you if this is the case. Obviously, if the
target disk is smaller a part of the source disk will not be copied, so this
is generally not a good idea.
disk cloning may take quite some time and both
disks will be active continuously. Because of this the disks may get hotter
than during normal operation.
Consider opening the PC casing and blowing in cool air using a fan placed at
approximately 1 meter from the PC.
If you do not apply
additional cooling, do not remove the case as this will disrupt the internal
as every read error is logged, the log
file may expand in size up to a point where no room is left on your
DiskPatch boot diskette. Consider disabling logging while cloning a disk
that is in really bad shape (read more on
how to do this here), or reduce the amount of info that is logged by
disabling 'log each disk read error' in
the options screen.
the write error threshold is set to '1'
for both steps of the cloning operation. If the prompt is displayed after a
write error has occurred, you can choose to disable the write error
threshold; after that, no more write error notifications will be displayed.
There are a number of configuration settings
that affect DiskPatch's behavior during cloning. Make sure these
settings are correct before starting the procedure.
The settings that affect cloning are:
||affects step 1
||affects step 2
|read error threshold
|write error threshold
|disk reset after error
|skip x sectors after error
|clone pass 2
|log each disk read error
affects logging only
|How to handle the resulting disk
after cloning is completed.
When the cloning procedure is complete you have a disk that should be
the mirror image of the original disk. How you handle this disk (the
clone) essentially depends on one thing: how well did the cloning go.
If there were many read errors and the original disk was having physical
problems you should be careful when accessing the clone. If you decide
to access the clone with Windows, the areas that were skipped because of
read errors may cause Windows to initiate repairs (chkdsk) and that will
ruin the work that the cloning procedure has done. Make sure nothing is
changed on the clone, don't let chkdsk or any other generic repair tool
access the disk. If normal access is not possible you should use a tool
to rescue files from the disk.
If the cloning went well and there were no read errors (for instance
when you clone a disk because you want a backup of the problem disk
before attempting a repair) you can now start working with the disk. You
can work either on the clone or on the original disk, though usually
it's best to work on the clone: when in a recovery situation you should
always attempt to minimize access to the original problem disk, to make
sure that things will not get worse. After all, if things go wrong when
working with the clone, you can make a new clone.
In this case too you should be careful when accessing the clone with
Windows; if volume damage exists Windows may initiate repairs and the
clone will be ruined.
Starting the clone procedure
Important: The disk that has been selected
in the 'select disk' screen is the source disk for the clone operation.
Select [Disk related tasks], [Clone]. Select the target disk
(the disk that should receive the clone). If the target disk is smaller than the
source disk DiskPatch will notify you.
You can then select the type of clone operation:
|| run step 1 of the
|| run step 1 and step 2 of the clone process.
|| run step 2 of the clone process (step 1 must have been
|| compare the cloned areas for the source- and target disk.
Next you must select the area to clone. Leave the suggested values to clone the
largest possible area. You can only select the area to clone when starting step
1; step 2 automatically uses the values that were entered when starting step 1.
Please note that the maximum size of the area that can be cloned is defined by
the smallest of the 2 disks involved.
The clone status screen:
The [Blocks skipped] count tells you how
many areas were encountered that have been skipped during step 1. The total
number of sectors skipped is determined by the 'block
skip size' and the number of areas skipped. For example: if 3 errors were
encountered, and the skip size is 1024 sectors then a total of 3072 sectors was
skipped; these would be cloned during step 2.
During step 2 the progress bar represents the amount of sectors that should be
copied for that step, not the total amount of sectors on the disk.
If DiskPatch encounters areas that can not be read during step 2 the 'estimated time
remaining' may increase dramatically. Once the bad areas are processed the
'estimated time remaining' will decrease again.