*SOS* urgent help needed.

[terminator2]

I totally agree with S.Haran
If situation demands then Imaging is must. Imaging should be performed on case to case basis .
In case disks are healthy ( SAS hard disks are rugged and chances of failure in first 3 years are far less).
If a target can be hit by a single bullet then why to deploy fleet of Apache helicopters backed by F-16 fighters ,that will be an overkill.

[northwind]

You've never heard of Murphy's law, have you?

[lcoughey]

I totally agree with S.Haran
If situation demands then Imaging is must. Imaging should be performed on case to case basis .
In case disks are healthy ( SAS hard disks are rugged and chances of failure in first 3 years are far less).
If a target can be hit by a single bullet then why to deploy fleet of Apache helicopters backed by F-16 fighters ,that will be an overkill.

And when you come over the mountain for the single target and are confronted with 1000 friends of the target pointing weapons back at you, you might wish that you were more prepared.

Assuming that a hard drive is healthy and won't fail is the number one reason that most projects get sent to labs for data recovery.

[S.Haran]

If any of the RAID members is unstable, it can fail on you and then you'll have some explaining to do.

We can never know how much life a drive has remaining. So I operate on the assumption that it has very little life left and can not survive the imaging process. If the drive fails while targeting the clients most important data at least I can tell the client X amount of data was saved then the drive died.

You must factor in that the imaging process itself is stressful and can cause drive failure in which case the client gets no data at all (unless one of you fine folks can bring it back to life). But I have never seen a case play out like this. It's rare but I do see RAID cases with failed drives but the drives have failed before I remote connect to the client.

Let me explain in brief my admittedly minority view of full disk imaging for RAID recovery. Imaging has two main drawbacks. It is costly in both time and materials. It places unneeded stress on a RAID drive by copying sectors the client does not care about (unallocated space / unneeded files)

The process I use is surgical and tries to minimize RAID drive stress. It attempts to first assemble the RAID and mount the filesystem. Prioritizing the recovery of the clients most important data first. But I do admit to using very small partial images when needed. The main problem I see in maybe 10% of cases is that the RAID will crash when it tries to read a file with bad sectors. But this can be worked around and a 99+% recovery can be achieved.

The benefits of this approach is speed, low cost, and it is remote recovery friendly. In simple RAID cases clients gaining access to their data in one hour is often possible.

Many of my clients are from underdeveloped parts of the world or do not have the financial means for a more traditional RAID recovery service.

My method is not for everyone. Strong Linux and RAID recovery skills are essential. My purpose in posting is to perhaps open your eyes to another way of skinning the RAID recovery cat.

[ddrecovery]

If any of the RAID members is unstable, it can fail on you and then you'll have some explaining to do.

We can never know how much life a drive has remaining. So I operate on the assumption that it has very little life left and can not survive the imaging process. If the drive fails while targeting the clients most important data at least I can tell the client X amount of data was saved then the drive died.

You must factor in that the imaging process itself is stressful and can cause drive failure in which case the client gets no data at all (unless one of you fine folks can bring it back to life). But I have never seen a case play out like this. It's rare but I do see RAID cases with failed drives but the drives have failed before I remote connect to the client.

Let me explain in brief my admittedly minority view of full disk imaging for RAID recovery. Imaging has two main drawbacks. It is costly in both time and materials. It places unneeded stress on a RAID drive by copying sectors the client does not care about (unallocated space / unneeded files)

The process I use is surgical and tries to minimize RAID drive stress. It attempts to first assemble the RAID and mount the filesystem. Prioritizing the recovery of the clients most important data first. But I do admit to using very small partial images when needed. The main problem I see in maybe 10% of cases is that the RAID will crash when it tries to read a file with bad sectors. But this can be worked around and a 99+% recovery can be achieved.

The benefits of this approach is speed, low cost, and it is remote recovery friendly. In simple RAID cases clients gaining access to their data in one hour is often possible.

Many of my clients are from underdeveloped parts of the world or do not have the financial means for a more traditional RAID recovery service.

My method is not for everyone. Strong Linux and RAID recovery skills are essential. My purpose in posting is to perhaps open your eyes to another way of skinning the RAID recovery cat.

Your response really is scary. It shows that you do not understand what imaging is, why it is done and how good it is for a failing drive. You do not understand modern data recovery methods such as connecting DDIs to R-Studio or PC3000 RAID to only image the clients data, not imaging every sector on every drive. Even if a full image is done on a failing member drive, it is far less stressful than saving off data via the file system.

[S.Haran]

I'll be kind and just say I'm glad you have a RAID recovery method that works for you. Does it also work well for remote RAID recovery? Because remote is all I do.

It's true I do not use your methods. Because in my experience RAID recovery can be achieved by other means. My Linux based method has been developed and refined over many years and is tried and true. We compete in a market place. Each client has to perform a cost/risk/benefit calculation. The market has a way of rewarding the recovery service that provides the best value.