HELP: Synology RAID5 array crashed

[jeffeuh]

Hi all,

Recently discovered my NAS DS1010+ in bad state:

 

Under HDD details, disk are in inconsistant state :

 

Don't know what to do to restore the volume.

Actually, cannot access to datas.

 

From CMD :

NAS1> cat /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4]
md2 : active raid5 sda3[7] sde3[5] sdd3[6] sdc3[2]
      7795173120 blocks super 1.1 level 5, 64k chunk, algorithm 2 [5/4] [U_UUU]

md1 : active raid1 sda2[0] sdc2[2] sdd2[3] sde2[4]
      2097088 blocks [5/4] [U_UUU]

md0 : active raid1 sda1[0] sdc1[4] sdd1[3] sde1[2]
      2490176 blocks [5/4] [U_UUU]

 

Tried to resync the volume but fail after some hours:

NAS1> mdadm --manage /dev/md2 --add /dev/sdb
mdadm: added /dev/sdb
NAS1> cat /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4]
md2 : active raid5 sdb[8] sda3[7] sde3[5] sdd3[6] sdc3[2]
      7795173120 blocks super 1.1 level 5, 64k chunk, algorithm 2 [5/4] [U_UUU]
      [>....................]  recovery =  0.0% (126976/1948793280) finish=255.7min speed=126976K/sec

md1 : active raid1 sda2[0] sdc2[2] sdd2[3] sde2[4]
      2097088 blocks [5/4] [U_UUU]

md0 : active raid1 sda1[0] sdc1[4] sdd1[3] sde1[2]
      2490176 blocks [5/4] [U_UUU]

I have tried this topic https://forum.synology.com/enu/viewtopic.php?f=77&t=146255&p=545651&hilit=crashed#p545651

But the volume stay in crashed and cannot access datas.

My first target is to access datas in degraded mode and backup them, but cannot ..

 

Any ideas ?

Many thx in advance for any help!

 

[knowledge-Spammer]

maybe a video on utube can help ?

https://www.youtube.com/results?search_query=Synology+RAID5+array+crashed

[jeffeuh]

Yes maybe.

The first video mention a tool "NAS Data Recovery" which may be a solution for me.

Thanks for your help, i will try it.

 

At the top of the top, I'm looking for an in place solution to recover the volume1.

Don't know if it's possible as is.

 

Thanks again. come back later with news.

[MrZeb]

Hi,

 

Why do you need to recover the data???

 

Its RAID 5 and you have only a bad HDD, simply replace the HDD and then let the Synology rebuild the RAID volume.

 

 

[straycat19]

3 hours ago, MrZeb said:

Why do you need to recover the data???

 

Its RAID 5 and you have only a bad HDD, simply replace the HDD and then let the Synology rebuild the RAID volume.

 

10 hours ago, jeffeuh said:

Under HDD details, disk are in inconsistant state

 

Normally I would agree with this but if you notice drive 2 says it is not initialized.  That is another problem.  That indicates that drive 2 also has a problem.  Replacing drive 3 will not fix the problem.  Drives that don't initialize indicate bad sectors, malware infection, corruption of the file allocation table, etc.  I don't know if this NAS can be recovered.  I have never seen one with two bad drives at the same time.  I have 6 NAS units at home (Drobo5N and 5N2 each with 5 10TB drives) and whenever a drive starts going bad I get a popup on the computer and a light on the front will turn from green to yellow or red.  The popup tells me which unit and which drive is bad and how to fix it, which is normally just to replace the drive.  You need to see if Synology has similar dashboard software you can install on your system to notify you the next time you have a problem.

[MrZeb]

5 minutes ago, straycat19 said:

 

 

Normally I would agree with this but if you notice drive 2 says it is not initialized.  That is another problem.  That indicates that drive 2 also has a problem.  Replacing drive 3 will not fix the problem.  Drives that don't initialize indicate bad sectors, malware infection, corruption of the file allocation table, etc.  I don't know if this NAS can be recovered.  I have never seen one with two bad drives at the same time.  I have 6 NAS units at home (Drobo5N and 5N2 each with 5 10TB drives) and whenever a drive starts going bad I get a popup on the computer and a light on the front will turn from green to yellow or red.  The popup tells me which unit and which drive is bad and how to fix it, which is normally just to replace the drive.  You need to see if Synology has similar dashboard software you can install on your system to notify you the next time you have a problem.

 

In this case both images show that the HDD2 isn't part of the RAID array (Volume 1), its simply a normal disk not initialized.

 

By default the Synology DSM software also show that type of warnings and also it uses beeps in case of problems unless the user disable that type of warning.

 

[jeffeuh]

At the start of RAID creation, 5 HDD are part of volume1.

I don't know the root cause, but the DISK2 has been "ejected" from the RAID and the DISK3 has no suffisiant health to rebuild the RAID. That's the issue.

This is not a "simple" case of HDD is failing and just replace it and rebuild the RAID.

I'm going to get all HDD to antother computer and try the solution of "NAS Data Recovery". For now it's my best way to hope recover data, during the time HDD3 is still alive.

 

I appreciate your, help.

If you have any suggestion, all are welcome.

 

Thanks again.

[LeetPirate]

13 hours ago, MrZeb said:

Hi,

 

Why do you need to recover the data???

 

Its RAID 5 and you have only a bad HDD, simply replace the HDD and then let the Synology rebuild the RAID volume.

 

 

 

This is correct, forget about disk 3 and what it says under SMART, just replace and initialise disk 2 and let it rebuild the RAID. If the system is preventing you due to the SMART status of disk 3 then you need a way to reset the smart parameters, for this you can try flashing the same version of firmware to the drive or newer firmware but data loss is a possibility, the second way is to use a jtag or serial flasher device to reset the smart parameters but I've never attempted or had a need to do this. Usually there is no reason SMART status would prevent the raid from rebuilding, if there is a way to disable smart on the nas you can safely do this until you rebuild your raid.

 

Your problem is not unusual, the reason the drive gets kicked from the RAID is because you are using desktop grade drives for NAS, this is not cool. Drives that are branded for NAS are so for a reason, the drive firmware needs to support TLER else when the drive finds an error it will not respond to the controller within the time limit and the raid controller will think the drive is non responsive and kick it out of the array. Some nutjobs on the Internet say that TLER isn't important and that Synology uses their own method of determining when to kick a drive out of the array, there may be some truth to this but I am sure TLER is still a major determining factor. Many people do the same thing you have done to save $2 and end up in the same situation or worse. If you recover from this situation you need to replace disk 2, 3, 4 with NAS grade drives one at a time, I recommend WD Red. I suggest that you first replace disk 2 with a new NAS grade drive and initialise it, let the raid rebuild itself.

[DKT27]

Topic moved to Technology Talk. Suits better here.

[macnavarra]

RAID 6 can sustain two drive failures, RAID 5 only 1.

[nonspin]

Before using any recovery software, go here:

https://www.cgsecurity.org/wiki/TestDisk

 

Download it, Read the Manual and Use it.

The tool is free and BY FAR the only one that is capable to restore/rebuild Arrays.

I've learned it the hard way..

[jeffeuh]

Hi all,

Come back with news.

My issue is the DISK1,3,4,5 are good in RAID5 array, BUT with DISK3 has bad SMART status, but make sense to be functional, instead the Synology cannot recover the RAID.

For an obscur reason that I can't determine, the DISK 2 is not initialized and due to the bad RAID5 status cannot be re-integrated into RAID array because of DISK3 status, and so array cannot be repaired.

 

So, regarding all feedback, I have discover an older post with a working full version of UFS Explorer Pro :

The latest working version is the 5.22.4 in 64bit.

 

This soft permits me to access on 4 DISK 1,3,4,5 and viewing the RAID5 array and copy them to another disks.

This is my first step, insure datas are collected and backed up.

Then, i will try the provided solution by @nonspin 

4 hours ago, nonspin said:

Before using any recovery software, go here:

https://www.cgsecurity.org/wiki/TestDisk

 

Download it, Read the Manual and Use it.

The tool is free and BY FAR the only one that is capable to restore/rebuild Arrays.

I've learned it the hard way..

 

If the soft is working well, I can restore the array and voilà !

If the soft is unable to recover the array, then I will destroy the RAID5 array and rebuild another with native SHR, to be able to add bigger disk with the same protection and extend my total volume size.

 

I really thanks all post to help me.

I Always hope there is a solution and I prefer to be optimist than thinking this is too late too early (as I have can read in some previous post).

Thanks again and "keep in nsane".

[nonspin]

if it doesnt work - you can revert the steps - That's why i recommended it.

.. Like i said, read the manual or watch the tutorial videos on that page.

 

The difference between testdisk and all the others is the method.

It's not trying to reconstruct your date byte-by-byte, it's operating

on the table above it - where it should.

If you'd know me, then you'd probably know that i'm not just handing out advice.