Please Help I am in Trouble

This is a good discussion, and from it, I or someone needs to write it up on the wiki.  There are some additional gotchas to discuss, depending on which route you go.

 

Regardless of which process you use, the safest approach is to disconnect ALL of the drives except the one you're working on at any given time.

 

This is an interesting option, does add safety in one area, but has hidden risks and disadvantages in other areas, particularly if the user is not running UnMENU.  The user is running v4.7, which made it hard to determine what the drive symbol was for a given drive, without UnMENU.  Not knowing it for sure or not realizing it had changed since the last boot, created the risk of processing the wrong drive, such as the parity drive.  It also creates the problem of not knowing where the physical drive is in the machine, to know for sure what drives to connect and disconnect.  The UnMENU MyMain screen has a feature to record those physical locations, and this would be VERY helpful right now!  And lastly, parity is invalidated by disconnections, and having to use the sd symbols.

 

Parity is also invalidated by the TestDisk approach, as nice as that option is.  The only option that preserves parity is keeping all drives just the way they are now, and using reiserfsck with the md symbols, which correspond to the disk numbers, no drive sd symbols needed, and no risks of the wrong drive being processed.

 

So parity protection is another factor in the decision of which of 3 methods to use -

* TestDisk - parity lost, must identify correct drive by model and serial, whether others are connected or not, must be able to identify additional drive symbol and be SURE it's empty, can still run reiserfsck afterward though

* reiserfsck of array drives - uses safe md symbols, no disconnections, parity is maintained, repair in place - so no TestDisk advantage

* reiserfsck on the only connected drive - must identify correct drive sd symbol (but then safer as can't run on wrong drive, if sure this is correct drive), parity is lost, repair in place - so no TestDisk advantage

 

One thing that is obvious but very important - either write down or print out the entire list of drives with their model and serial numbers and disk numbers.  You will be adding notes to it as you progress.  Then it would also be useful to add the physical location of each drive, and/or use the MyMain feature.

 

I considered discussing the option of upgrading to v6, partly because it's safer, and partly because sd symbols are always displayed, but decided not to for now.

 

I do think some caution is warranted though, because we don't yet know why a group of drives failed to mount, and so it could happen again.  When a group fails together, then you worry about controller failure or power issues.  Really wish we had the syslog from when the trouble occurred.  If the added drives caused increased power demands, then disconnections are a good choice, and a PSU upgrade!

... I considered discussing the option of upgrading to v6, partly because it's safer, and partly because sd symbols are always displayed, but decided not to for now.

 

??  I don't think this is a viable option at this stage, as the data should be recovered first.  Upgrading could introduce several common v4.7 => newer version pitfalls ... HPA(s), plugins that don't work, etc.    I think the key focus now should be on data recovery.

 

 

"... Parity is also invalidated by the TestDisk approach, as nice as that option is." ==>  Why?  TestDisk makes NO changes to the disk being recovered.

 

I certainly agree that step #1 should be to documents the model/serial # of all drives and their current assignments.  Fortunately, this is already done in Reply #3 

 

While it's not actually necessary to disconnect the drives, it's a good "safety play" that will guarantee not selecting an incorrect drive to work on.  As for invalidating parity => I'd think that at this stage parity isn't useful anyway, since any drive rebuild using the parity info is just going to rebuild the formatted drive.    I suppose if you use the ReiserFsck alternative with the md drive references, you can indeed preserve parity ... but I'd consider doing so a relatively low priority part of this process.

 

I also agree that there could be controller and/or power issues at work here -- yet another reason to only work on one drive at a time.    Note this doesn't even need to be in the same system ... you could connect one drive at a time to another system, and do the recovery there.

 

 

 

Firstly, Gary,RobJ,Itimpi

 

Thank you for your responses to my desperate situation I am truly humbled by the content of your replies which clearly you have taken considerable time to work out.

 

It is clear to me the work involved here for me is considerable,but the stress of the current situation is a great leveler I am prepared to pick up the mantle just on that alone!

 

I have never undertaken such a massive piece of work before so the learning curve will be a steep one. Its clear to me from your replies that an important issue is one of patience, taking, my time, and getting it right, to acquire the maximum results.

 

I do of course have questions that I am unsure of, I do run Unmenu so hopefully that will be of assistance.

 

1. I assume that drives cannot be removed from the machine whilst a drive is being examined? I would prefer not to disconnect drives from a cable perspective if that could be avoided.

 

2. What is the importance of maintaining parity drive, although i do have one drive currently left with 80% information on it. I am not sure if that was affected by the format I did note activity lights on during the process.

 

3. Are the drives renamed during the process? so they do not look the same.

 

4. I dont have a CD facility on the box, I do have a CD on my computer, but I am unsure of the process of re-booting the box using a CD with TestDrive. So your help/views would be welcome.

 

5. I noted in the correspondence that the provision of another drive is required for TestDrive to write to. Thereby hangs the original issue giving rise to the current predicament. I have two precleared drives in the box that 4.7 is currently not seeing to format. So I would of course need to sort that issue out before this one, so any help there would be great.

 

6.At the moment I see that there is an advantage reiserfsck though I have no experience of using it, and for that matter I dont even know where it sits!!.Am I right in thinking that this process does not require and additional drive? and its replaces what it finds on the original drive?. I understand the pains taking aspect of me searching through.Would be able to search material using Windows Tower Drive 1 or has this got to be through Putty or something?

 

7. Am I still able to use the existing drive that was not affected by the disaster? without prejudicing anything?

 

8. Finally, I have not emailed Tom as yet, your views?

 

Sorry about the  time lag im in UK

 

Peter

 

... an important issue is one of patience, taking, my time, and getting it right, to acquire the maximum results.

 

Not just "an important issue" ... but THE important issue.    This will absolutely require significant patience and perseverance.

 

Note that it's very easy to add a CD drive to your system => just use an external USB unit.    As long as your BIOS supports booting from a USB device, that will let you do what you need to do.

 

r.e. Rieserfsck ... yes, it's already included in Linux, so it's just a matter of running the proper commands.    As long as you're VERY, VERY CAREFUL about the command and the drive reference each time, you can safely use it one drive at a time without disconnecting anything ... and as Rob noted, it will (as long as you use the md references, NOT the sd) parity will be preserved, so if you did have a drive fail during this process you could replace it and the system should correctly rebuild the replacement.

 

The key thing right now is to be CERTAIN that you understand exactly what each command you use is going to do.  I would NOT do ANYTHING on the server until you've done this recovery.    If anything resulted in a write to one of the incorrectly formatted drives, your ability to recover data on that drive would be significantly degraded.

 

 

 

... I considered discussing the option of upgrading to v6, partly because it's safer, and partly because sd symbols are always displayed, but decided not to for now.

??  I don't think this is a viable option at this stage, as the data should be recovered first.  Upgrading could introduce several common v4.7 => newer version pitfalls ... HPA(s), plugins that don't work, etc.    I think the key focus now should be on data recovery.

Agreed.  Plus there's the learning curve of v6, not a good time for that right now!  Fortunately, when the recovery is complete, you will be in a good position to upgrade then, a 'clean' install, then do the parity rebuild (if needed) and New Permissions (and learn about all the new stuff).

 

"... Parity is also invalidated by the TestDisk approach, as nice as that option is." ==>  Why?  TestDisk makes NO changes to the disk being recovered.

While that's true, that's only the first phase.  You still have to deal with the old drive, and you have to get the data back into the array.  Both removing the old drive and adding the new drive kills parity.  There IS a way to preserve parity, format the old drive (may already be!), manually mount the temporary drive, and copy the data back to the old drive, but that involves extra steps and much more time.  But another problem, the drive with the TestDisk recovered data is not part of the array, and therefore hard for him to examine and move files around.  No easy way to do that without killing parity.  Plus, the new drive would have to be a brand new drive not already in the array!  He's probably going to want to use one of his new drives, which are already in the array I think.  I could be wrong here.

 

TopTrainer: as Gary said, reiserfsck is built in, just follow the instructions in the Check_Disk_Filesystems#Drives_formatted_with_ReiserFS_using_unRAID_v4 section.

 

TestDisk is a 'live CD' type tool, a tool built onto its own Linux OS just like unRAID is, and can be burned onto a bootable CD or even (with another tool) installed on a bootable flash drive.  In fact if we had the right Linux guru here, they could probably give us the steps to install it on the unRAID flash drive and add it to the unRAID boot menu!  I hope someone does, and writes the procedure up for us.  To use TestDisk on a CD, you will need to add a CD drive to the unRAID system, or set up another system with CD drive and add the problem drive plus another empty drive to it.  If either of those choices is available, this is still a good option.  I personally don't think it will be feasible to keep the parity protection this way, but I'm not sure how important that is at this point.  I'm sorry to sound ambivalent on the choices, but I don't see a clear winner yet, rather you make the choice based on what you feel most comfortable with (or LEAST uncomfortable with!).

As you can tell, there are a lot of potential ways to approach this, but I agree with Rob's final point => STUDY the alternatives; and take the approach you feel the most comfortable with.    I won't echo Rob's "... LEAST uncomfortable with ..." ==> if you're uncomfortable you should STUDY some more and perhaps even practice on another system until you ARE reasonably comfortable that you understand the process you're going to follow. 

 

Which approach would I take ??    I'd use TestDisk, primarily because you can do that and STILL try Reiserfsck later if it isn't successful; whereas once you run Reiserfsck the disk will be modified and not recoverable via other tools (like TestDisk).

 

I value both of your opinions, I knew there was not going to be any easy return from this situation,so I fully understand the task ahead.

The finer points of getting started are what I want to finalise which is what I think we are currently discussing, yes there is some technical issues that I need to get my head around and understand the importance of strictly following the process whatever that might be.

 

I can see benefits of TestDisk and also see Gary's point about it allows for a second option too. I guess being honest I am not comfortable about booting the server from the a CD for the reasons stated. However Rob speaks of another option  adding TestDisk to the bootmenu. That would be something to consider if available.

 

I seem to have got a bit overloaded at the moment, regarding parity wont this need to be completely rebuilt? I am just looking at conversation around preserving it surely that wont be possible will it?

 

So to clarify for me please, Rieserfsck will recover data in situe on the single drive its reading?

 

TestDisk will recover data  but requires another disk to write it to. And will also allow for a second shot with Rieserfsck?

 

At the moment guys im feeling a bit more comfortable with Rieserfsck given its part of the system. And I do run UnMenu. I will construct a map from UnMenu, and start to read all of the material you have pointed me towards ahead of the task.

 

I do need to discuss the issues with you around the two new drives that are currently in the server but are not in the array because the format process did not touch them. Which of course gave rise to the current situation.

I am unsure why they are not listed in the device drop downs either? I would like to add them to the array before I start the recovery process as I fear some sort of repeat and do not want to damage any recovered work. So I would appreciate your advice on the way forward with that issue.

Again I am very grateful for the benefit of all your experience and advice and the time you given to writing it. I am comforted by the fact I can still consult you during the process but also realize a successful return will make me somewhat of an expert too!!

 

Peter

 

 

I do need to discuss the issues with you around the two new drives that are currently in the server but are not in the array because the format process did not touch them. Which of course gave rise to the current situation.

I am unsure why they are not listed in the device drop downs either? I would like to add them to the array before I start the recovery process as I fear some sort of repeat and do not want to damage any recovered work. So I would appreciate your advice on the way forward with that issue.

 

Please see Capturing your syslog, and attach your syslog, so we can determine what is going on with the drives.

Since this trouble started when trying to increase your drive count, what is the exact model of your power supply?

If you find you want to run reiserfsck, let me know and I can support. I have done it myself, and walked a number of users through the process several times. When it works, which has been most of the time, the results have been pretty good, although some data is lost and there can be some moving of data around.

If you find you want to run reiserfsck, let me know and I can support. I have done it myself, and walked a number of users through the process several times. When it works, which has been most of the time, the results have been pretty good, although some data is lost and there can be some moving of data around.

I hadn't mentioned you directly, but was expecting him to find the links to your experiences, and shortly become VERY familiar with your writeups!   

A few more thoughts ...

 

=> r.e. preserving parity.    Parity SHOULD be just fine at the moment ... it would have been updated as the drives were formatted, so any failed drive could in fact be restored to a replacement and you could still do recovery from the replacement drive.

 

HOWEVER ... it's important to remember that SOMETHING happened to cause all this => it could be inadequate power (as trurl just implied);  a bad SATA controller port;  bad memory; or any of several other potential issues.

 

One thing I'd do NOW ... BEFORE anything else ... is to boot the server and select MemTest instead of UnRAID.    Let it run for several hours (at least 3-4 full cycles) ... or even overnight.    Memory is perhaps the easiest thing to eliminate as an issue (or at least get high confidence that it's not the issue).

 

You shouldn't be concerned about booting to a TestDisk CD ... doing that is NO different than booting to the UnRAID flash drive and selected TestDisk as an option (if it was added to it).    In either case, at that point you'd be running TestDisk -- NOT UnRAID.    It's effectively no different than booting to a MemTest CD vs. just selecting MemTest from the UnRAID flash drive's boot menu.

 

Since you're leaning towards using Reiserfsck, you want to be sure you've eliminated memory and power as issues before you start doing the recovery.    Memory is easy to eliminate (per MemTest) ... but power's a bit trickier to be certain about without extensive test gear.    The easiest way to have high confidence it's not power is to replace the power supply with a high quality unit that you KNOW has plenty of extra "headroom".

 

As for the mechanics of recovering with ReiserFsck =>  IF you use the correct MD disk references while doing the recovery, then parity will be preserved throughout the entire process ... so even if a disk fails you'll be able to replace it and continue the recovery.    You're correct that it writes the corrections directly to the disk being recovered -- and as long as it's using the MD references, UnRAID will be updating parity accordingly as it proceeds.      The only downside of doing this is that if it doesn't succeed, there's no alternative recovery you can try, as the disk will have already been modified.  Note that data recovery professionals will tell you that the #1 rule of data recovery is NEVER modify the disk you're recovering from (which is why TestDisk doesn't modify the disk at all).

 

r.e. adding the 2 new disks you have before doing the recoveries =>  you should be able to do this IF the disks show up on the drop-down lists for unused slots while the array is stopped  (and if they're pre-cleared they'll add very quickly).  But remember that this is what went wrong in the first place ... so be sure you've tested memory first AND are fairly confident that you don't have a power issue before you try this again.    Note that doing this won't (or I should say SHOULDN'T) impact the validity of your parity drive ... it will be appropriately updated to reflect the new drives and should still be valid.

 

Glad Brian (bjp) chimed in -- he is indeed VERY good at guiding you through the ReiserFSCK process.

 

Since this trouble started when trying to increase your drive count, what is the exact model of your power supply?

trurl thanks for your response and the power supply = Corsair SX 600.

 

If you find you want to run reiserfsck, let me know and I can support. I have done it myself, and walked a number of users through the process several times. When it works, which has been most of the time, the results have been pretty good, although some data is lost and there can be some moving of data around.

Brian, thank you this option is currently holding my interest, though I do understand what Gary states regarding its use being a somewhat final action. I will of course see your assistance when I do make a start.

I do need to discuss the issues with you around the two new drives that are currently in the server but are not in the array because the format process did not touch them. Which of course gave rise to the current situation.

I am unsure why they are not listed in the device drop downs either? I would like to add them to the array before I start the recovery process as I fear some sort of repeat and do not want to damage any recovered work. So I would appreciate your advice on the way forward with that issue.

 

 

Please see Capturing your syslog, and attach your syslog, so we can determine what is going on with the drives.

Rob I will attach a sys log but i should point out the server is currently off so it will be the log on a reboot, I hope that will be ok.

 

 

A few more thoughts ...

 

 

 

HOWEVER ... it's important to remember that SOMETHING happened to cause all this => it could be inadequate power (as trurl just implied);  a bad SATA controller port;  bad memory; or any of several other potential issues.

 

One thing I'd do NOW ... BEFORE anything else ... is to boot the server and select MemTest instead of UnRAID.    Let it run for several hours (at least 3-4 full cycles) ... or even overnight.    Memory is perhaps the easiest thing to eliminate as an issue (or at least get high confidence that it's not the issue). will try the mem test, i have never run that before.

 

 

 

Since you're leaning towards using Reiserfsck, you want to be sure you've eliminated memory and power as issues before you start doing the recovery.    Memory is easy to eliminate (per MemTest) ... but power's a bit trickier to be certain about without extensive test gear.    The easiest way to have high confidence it's not power is to replace the power supply with a high quality unit that you KNOW has plenty of extra "headroom".  current =Corsair SX 600

 

 

 

r.e. adding the 2 new disks you have before doing the recoveries =>  you should be able to do this IF the disks show up on the drop-down lists for unused slots while the array is stopped  (and if they're pre-cleared they'll add very quickly).  But remember that this is what went wrong in the first place ... so be sure you've tested memory first AND are fairly confident that you don't have a power issue before you try this again.    Note that doing this won't (or I should say SHOULDN'T) impact the validity of your parity drive ... it will be appropriately updated to reflect the new drives and should still be valid.

Gary I precleared both these drives which went fine, but when I stopped the array and then re-started it, neither were shown in the drop down menu, I tried this several times. I am unclear as to why they should not be listed. i will be re-booted again tonight to see what happens.

All

 

syslog  attached. UPDATE the two new drives have now been formatted and added to the array

 

Peter

syslog-2015-09-02.txt

Very good -- that also tends to confirm that you have ample power.  You should have enough with 600w, although the CX series (which I presume is what you meant when you said "SX 600") is the bottom of Corsair's line, and DOES have more issues than their other units (the TX, HX, and AX are much better).

 

I'd run the memory test I suggested before you actually start doing any recover -- and then you'll be ready to start the one-disk-at-a-time recovery attempts using ReiseerFsck with Brian's guidance.

 

Syslog shows 16 drives connected, and more SATA ports available!  The array has 14 drives, 12 data and a parity and cache drive.

 

Your SAS drive is using an early version of the SAS support, the mvsas+scst software, which was rather young then.  It gave me no confidence at all, when examining syslogs with it, and Tom dropped it almost immediately in the early v5 betas.  I used to try to push v4.7 users to upgrade to any v5 beta version if I saw the mvsas+scst stuff, as I felt that even if v5 was in beta then, it was still better and safer SAS support.  So once the recovery is far enough along, you will need to consider upgrading to v6, better and safer SAS support for your hardware.

 

Since it's working fine at the moment, there are no issues visible in the syslog, and no evidence as to what happened previously.  If anything wrong happens again, capture that syslog!

Rob thanks for reviewing the sys log I am pleased things are not as bad there as perhaps we thought. I don't know anything

about mvsas+scst but if you are saying you have concerns around it that is good enough for me and when the rebuild process is complete I would like to upgrade to Pro V6 so will be seeking the appropriate advice and guidance to do that.

 

Gary, your observations on the power supply are also noted and based upon your views and current recommendations for V6 I will also replace the power supply.

 

It is my intention to start the recovery process as soon as practicable and I await contact from Brian with the initial guidance and instructions for commencement. I am fortunate to run UnMenu and I will be mapping the drives using it, I also have the luxury of large dual monitors so i can actually have UnMenu open as I do things to support and quality control my actions.

 

I will of course update with my progress as and when it is achieved at the moment I have no real idea what the timescale of each drive looks like. I am extremely grateful for your continued support and advice.

 

Rob thanks for reviewing the sys log I am pleased things are not as bad there as perhaps we thought. I don't know anything

about mvsas+scst but if you are saying you have concerns around it that is good enough for me and when the rebuild process is complete I would like to upgrade to Pro V6 so will be seeking the appropriate advice and guidance to do that.

 

Gary, your observations on the power supply are also noted and based upon your views and current recommendations for V6 I will also replace the power supply.

 

It is my intention to start the recovery process as soon as practicable and I await contact from Brian with the initial guidance and instructions for commencement. I am fortunate to run UnMenu and I will be mapping the drives using it, I also have the luxury of large dual monitors so i can actually have UnMenu open as I do things to support and quality control my actions.

 

I will of course update with my progress as and when it is achieved at the moment I have no real idea what the timescale of each drive looks like. I am extremely grateful for your continued support and advice.

 

I am traveling and will post some information tomorrow.

Please review THIS THREAD for another user's experience and support from me and Tom.

 

Post back once you have reviewed and are ready to start.

Brian, and Everyone else currently involved in helping to rescue me from this disaster;

 

Thanks for the link, sorry I have been a while responding, but I have been reading it over and over!!!

to say i am scared to death, quite frankly is an understatement!!! Whilst realize your undertaking on that particular disaster differs greatly from my situation.

At this moment in time I feel like I am on a tightrope stretched out across the Grand Canyon and about to walk across having never tried it before!!. I think, I have underestimated what is required here, and I don't even know if I possess the ability to do most of this.(massive learning curve ahead)

 

The link refers to one drive, suddenly I thought my god there is 11 2tb drives here and each one fairly full. I am calm in that I am not in a panic situation trying to prevent something happening that would likely wipe everything out. The server is on just sitting quietly and nothing has been written to the formatted drives.

 

Not surprisingly I have some questions which clearly I foolishly had not considered;

 

1. Will I be able to use Putty for this? I used that to preclear the drives, as I have never ever used the console not even sure how to find it. I was hoping to copy and paste commands.

   

 

2. I had not envisaged that I would need to view the recovered data somewhere else and validate, and copy it to another drive. I am not sure quite how I will do that. Is there something I need to buy?

                     

3. will I be able to view the system still through UnMenu whilst the drive is being worked?

 

I want to try to avoid the stressful cliffhangers as I think I am possibly a lot older than the star in the thread

and keeping my stress levels low will be beneficial.

 

 

Peter

 

Don't get stressed -- this process simply requires a LOT of patience and is far more "wait time" than actually "doing anything" time.

 

Essentially, you simply have to type 11 commands on the Linux command line -- one at a time and MANY hours apart.  Each command will run for a long time.

 

THEN you will likely have to look at a bunch of recovered files and rename them appropriately.

 

If you do this via the MD references (as Brian will ensure you do) then parity will be maintained throughout the entire process ... and you won't have to copy the files to any other location in the process  [ReiserFSCK will recover them "in place" on the disk.

 

r.e. Putty => I'm not a "Linux guy", so take this answer with a grain of salt ... but my understanding is that Putty will, like all Telnet clients, terminate a session if you close Putty.    There are apparently a couple of command parameters you can append so this won't happen, but then you can't reconnect to the same session.    If you must do this remotely, I'd use Screen to connect to the server; as this can easily be closed and then later reconnected to the same session.    But ideally I'd do this on a display and keyboard connected to the actual UnRAID server.

 

r.e. UnMenu => you should still be able to view the UnMenu GUI while you're using ReiserFSCK

 

r.e. age  :)   [some of us have a few years under our belt as well ... although Brian is indeed a youngster  ]

 

 

1. Will I be able to use Putty for this? I used that to preclear the drives, as I have never ever used the console not even sure how to find it. I was hoping to copy and paste commands.

Yes you will need a telnet session, and if using Windows then Putty is the favoured way of running one.  This an alternative to a physical screen/keyboard.  You must not close a Putty window while running reiserfsck as (unless you have taken special action) this would terminate the reiserfsck.

2. I had not envisaged that I would need to view the recovered data somewhere else and validate, and copy it to another drive. I am not sure quite how I will do that. Is there something I need to buy?

Another drive is only required if you go down the TestDisk (or equivalent) route as in that case you are leaving the source disk untouched and putting the files found onto the other disk.  The upside of this approach is that you can be sure no changes are made to the source disk while attempting recovery in case something goes wrong and you need to try again.  The downside is needing another drive to put recovered files onto.

 

If you go down the reiserfsck then the fixes are 'in-situ' and another disk is not required.  The reiserfsck modifies the disk while it is running attempting to fix it. The upside of the reiserfsck method is that it can be done on the unRAID server and after completing a disk you can examine the array to see how successful recovery has been before attempting another disk.    The downside of the reiserfsck approach is that if it goes wrong you have already modified the disk so there is no alternative recovery method available.

                    3. will I be able to view the system still through UnMenu whilst the drive is being worked?

While you are working on a disk the array will be in 'maintenance' mode.  You will be able to run the GUI (or unMenu) but the data disks will not be mounted so you cannot see their contents.

Brian, and Everyone else currently involved in helping to rescue me from this disaster;

 

Thanks for the link, sorry I have been a while responding, but I have been reading it over and over!!!

to say i am scared to death, quite frankly is an understatement!!! Whilst realize your undertaking on that particular disaster differs greatly from my situation.

At this moment in time I feel like I am on a tightrope stretched out across the Grand Canyon and about to walk across having never tried it before!!. I think, I have underestimated what is required here, and I don't even know if I possess the ability to do most of this.(massive learning curve ahead)

 

Stress is is a useful tool to keep you from making careless mistakes. Just don't let it keep you from doing the needful.

 

The link refers to one drive, suddenly I thought my god there is 11 2tb drives here and each one fairly full. I am calm in that I am not in a panic situation trying to prevent something happening that would likely wipe everything out. The server is on just sitting quietly and nothing has been written to the formatted drives.

 

The stress will (hopefully) enormously reduce after doing the first drive. Assuming it works as expected, the other drives will be much less stressful as you will have confidence in a positive outcome.

 

Not surprisingly I have some questions which clearly I foolishly had not considered;

 

1. Will I be able to use Putty for this? I used that to preclear the drives, as I have never ever used the console not even sure how to find it. I was hoping to copy and paste commands.

 

Yes, inside a screen session. We can work on setting this up if it is not already on your system.

   

2. I had not envisaged that I would need to view the recovered data somewhere else and validate, and copy it to another drive. I am not sure quite how I will do that. Is there something I need to buy?

 

You may indeed need a second drive. After doing my recovery, I copied everything to a newly formatted disk and reformatted the original, which had weirdly named files and directories that did not try to clean up. But you should only need one - and after doing all the drives, you should have one empty drive left over.

                     

3. will I be able to view the system still through UnMenu whilst the drive is being worked?

 

Yes

 

I want to try to avoid the stressful cliffhangers as I think I am possibly a lot older than the star in the thread

and keeping my stress levels low will be beneficial.

 

Feels good to be a youngster again. Thanks Gary, you made this quinquagenarian's day!