I've actually decided to try recover the data. I have a lot of irrecoverable stuff on there, and even the recoverable stuff will take months or years to build back up. If I take it to a forensic analyst, what are the possible outcomes? I mean, what are the possibilities that would get my zfs volume back? Physical disk recovery? I assume partial data recovery won't be of any use to me. Basically I want to know that if he calls me up and tells me what he can do, I'll be able to tell him yes or no.
-
Important Announcement for the TrueNAS Community.
The TrueNAS Community has now been moved. This forum will now become READ-ONLY for historical purposes. Please feel free to join us on the new TrueNAS Community Forums
Replacing a failed disk. Dead drive not showing anywhere
- Thread starter Zarzob
- Start date
- Status
I've actually decided to try recover the data. I have a lot of irrecoverable stuff on there, and even the recoverable stuff will take months or years to build back up. If I take it to a forensic analyst, what are the possible outcomes? I mean, what are the possibilities that would get my zfs volume back? Physical disk recovery? I assume partial data recovery won't be of any use to me. Basically I want to know that if he calls me up and tells me what he can do, I'll be able to tell him yes or no.
You don't want to go to a forensic analyst. You want someone to simply recover the hard drive back to an operable state and return it; or return the drive to a state that they can do a disk copy to a new disk to return to you. Once that's done you can mount and recover the pool yourself.
Any forensic analysis is going to be ungodly expensive and you won't be able to afford it(the tool to recover ZFS must be custom written for each situation).
Anyone claiming to offer "forensic analysis" that isn't ungodly expensive knows that they just have to repair the broken disk and mount the pool so they're asking you to pay for a service you don't need but they may be able to sucker you into paying. ;)
Any forensic analysis is going to be ungodly expensive and you won't be able to afford it(the tool to recover ZFS must be custom written for each situation).
Anyone claiming to offer "forensic analysis" that isn't ungodly expensive knows that they just have to repair the broken disk and mount the pool so they're asking you to pay for a service you don't need but they may be able to sucker you into paying. ;)
- Joined
- May 28, 2011
- Messages
- 10,994
So I take it that you didn't have any backup of what you considered valuable data?
solarisguy
Guru
- Joined
- Apr 4, 2014
- Messages
- 1,125
Forensic analyst? A person working in digital forensics needs to be somehow certified, so the evidence is admissible in the courts. You only need a recovery specialist, their forensic analyst title only possibly boosting their credentials. In the USA, as mentioned by cyberjock, there could be a substantial cost difference. I am aware of a European country, where a global recovery company has two local Internet domains, one for the legal work and one for the work that would not be entering the legal system. The pricing is different, the work is being done by the same team.
You are hoping for all the data intact and accessible, after the electronic board and possibly mechanical parts are replaced, as that means your pool would be back. A relatively simple procedure with a known outcome that seems to correspond to the quote you had received.
Partial data recovery would mean that some part, possibly very small, was destroyed and is nonrecoverable. Partial data recovery would happen if the heads landed on the platters and physically damaged the surface. Often there is a need to recover at least some part of what was destroyed and it could be that someone would be looking under a microscope (a gross oversimplification) and reading bits... Connection to Sherlock Holmes is immediate = a forensic analyst etc. Some data would be lost.
You are hoping for all the data intact and accessible, after the electronic board and possibly mechanical parts are replaced, as that means your pool would be back. A relatively simple procedure with a known outcome that seems to correspond to the quote you had received.
Partial data recovery would mean that some part, possibly very small, was destroyed and is nonrecoverable. Partial data recovery would happen if the heads landed on the platters and physically damaged the surface. Often there is a need to recover at least some part of what was destroyed and it could be that someone would be looking under a microscope (a gross oversimplification) and reading bits... Connection to Sherlock Holmes is immediate = a forensic analyst etc. Some data would be lost.
- Joined
- May 28, 2011
- Messages
- 10,994
If you can hear the drive motor going and no squeaking (bad noises), you might have a chance with just the board replacement. I think there is some good talk about drive recovery but the cost is probably too high when you don't even recall what was on the drive for sure.
- Joined
- May 28, 2011
- Messages
- 10,994
The good thing about starting over this time is you know how you use your system and can build the pool properly. If you wanted a RAIDZ2, now would be the time to do it. I don't recall if you stated that you are using ECC RAM or not. If not then you might consider making that change.
I'm using 16GB Non-ECC ram.
I have a few questions:
- If in the future I want to upgrade to ECC ram (a bit expensive for me right now), will I run into any issues? Do I need to do anything in particular or can I just switch out the RAM?
- I now have 8x 3TB hard drives. I wanted to set it up as RAIDZ3, but FreeNAS told me it was non-optimal. I read up quite a few posts about the amount of drives you should be using on the different RAID configs, but still can't fully understand why. What exactly is the difference I can see? In Cyberjock's FreeNAS guide it says it was only a thumbrule, and can be ignored. It's just a personal file server, so other than using the Transmission or Plex Media Server plugins, there will be very few times when I will be doing more than one simultaneous read/write. Will I notice a difference in performance or should I just go for RAIDZ3 for maximum redundancy?
- I remember reading somewhere that replacing a drive in RAIDZ1 configuration, you will no longer have any redundancy at all. If in the future I replace a failed hard drive, will it still have the same amount of drive redundancy as before after resilvering? i.e. if I start a RAIDZ3 with 3 drive redundancy, if I replace one of the drives will it still have 3 drives worth of redundancy, or two? Likewise with RAIDZ2, will I still have 2 drives of redundancy after replacing one?
- In the "Advanced Disk Management" section of the disk setup, if I wasn't worried about power usage, what would be the optimal setting to maximise the life of the disks (is there anything else I can do to maximise the life of disk)? In terms of preserving the heads and keeping the temperature low, or anything else, if any? I'm using Seagate Barracuda's if it makes a difference.
I have a few questions:
- If in the future I want to upgrade to ECC ram (a bit expensive for me right now), will I run into any issues? Do I need to do anything in particular or can I just switch out the RAM?
- I now have 8x 3TB hard drives. I wanted to set it up as RAIDZ3, but FreeNAS told me it was non-optimal. I read up quite a few posts about the amount of drives you should be using on the different RAID configs, but still can't fully understand why. What exactly is the difference I can see? In Cyberjock's FreeNAS guide it says it was only a thumbrule, and can be ignored. It's just a personal file server, so other than using the Transmission or Plex Media Server plugins, there will be very few times when I will be doing more than one simultaneous read/write. Will I notice a difference in performance or should I just go for RAIDZ3 for maximum redundancy?
- I remember reading somewhere that replacing a drive in RAIDZ1 configuration, you will no longer have any redundancy at all. If in the future I replace a failed hard drive, will it still have the same amount of drive redundancy as before after resilvering? i.e. if I start a RAIDZ3 with 3 drive redundancy, if I replace one of the drives will it still have 3 drives worth of redundancy, or two? Likewise with RAIDZ2, will I still have 2 drives of redundancy after replacing one?
- In the "Advanced Disk Management" section of the disk setup, if I wasn't worried about power usage, what would be the optimal setting to maximise the life of the disks (is there anything else I can do to maximise the life of disk)? In terms of preserving the heads and keeping the temperature low, or anything else, if any? I'm using Seagate Barracuda's if it makes a difference.
- Joined
- May 28, 2011
- Messages
- 10,994
Quite a few questions:
RAIDZ being optimal is really something for an busy office environment, non-optimal is fine for everything else.
The higher the RAIDZ level, the slower your pool will be. It may not be much slower, or even noticeable, but it will be slower.
When replacing a failed drive with a RAIDZ1, until your new drive is resilvered you will have no redundancy, however once it is resilvered you will have the same 1 drive redundancy you started with.
For a home system I would recommend a RAIDZ2 system.
To maximize the life of your drives, in my opinion you should leave the drives spinning because constant spinning up and down will kill a drive, and keep the temperatures below 40C. Above 40C and the thought is your drive lifespan will be reduced. There are a lot of threads on this topic.
As for ECC RAM, using ZFS you should be using ECC RAM. All you need to do is verify your MB supports it, buy the correct compatible RAM, swap it out and possibly configure the BIOS to recognize the ECC RAM installed. That is about it. Sounds simple and it "should" be.
Good Luck.
RAIDZ being optimal is really something for an busy office environment, non-optimal is fine for everything else.
The higher the RAIDZ level, the slower your pool will be. It may not be much slower, or even noticeable, but it will be slower.
When replacing a failed drive with a RAIDZ1, until your new drive is resilvered you will have no redundancy, however once it is resilvered you will have the same 1 drive redundancy you started with.
For a home system I would recommend a RAIDZ2 system.
To maximize the life of your drives, in my opinion you should leave the drives spinning because constant spinning up and down will kill a drive, and keep the temperatures below 40C. Above 40C and the thought is your drive lifespan will be reduced. There are a lot of threads on this topic.
As for ECC RAM, using ZFS you should be using ECC RAM. All you need to do is verify your MB supports it, buy the correct compatible RAM, swap it out and possibly configure the BIOS to recognize the ECC RAM installed. That is about it. Sounds simple and it "should" be.
Good Luck.
- Joined
- Feb 15, 2014
- Messages
- 20,194
If the motherboard is compatible and if the CPU is compatible, then all you have to do is remove the old RAM and add the ECC RAM.
solarisguy
Guru
- Joined
- Apr 4, 2014
- Messages
- 1,125
There are motherboards like Ericloewe has just described, however they are not the most popular ones. Please, please check whether your motherboard maker claims ECC support or just support for ECC RAM**** For recent Intel CPU, it is rather easy: the motherboard would need to have C22x or C20x chipset.Ericloewe said:
**** There are quite a few motherboards, that you can insert into them either ECC or non-ECC RAM, but no ECC support is offered = ECC memory lines are ignored.
- Status
