Expand storage pool nightmare

[FAsnakes]

Everything was working wonderfully for years and I’m pretty sure I screwed this up..

Here’s my setup:
Synology DS920+ running DSM 7.1.1 Update 4

• Drive 1 18TB Ironwolf NAS HD
• Drive 2 18TB Ironwolf NAS HD
• Drive 3 12TB Ironwolf NAS HD
• Drive 4 12TB Ironwolf NAS HD

Running SHR 1 drive fault tolerance all under 1 storage pool.

Here’s what I did:

1. Replaced drive 3 with a 20TB Ironwolf HD and it was not detected.
2. I then took the 20TB drive that wasn’t detected and put it in Drive 4.

This is when everything broke. In a panic I removed the 20TB drive and put everything back but the damage was already done.

On drive 3 I got a message saying ‘the system removed this drive from the storage pool. To continue using this drive, please go to Action > Deactivate Drive to deactivate the drive first. Afterward, you can either restart you Synology NAS or remove and reinsert the drive.’

Currently, in storage manager I’m seeing:

• Storage pool 1 degraded
• Drive 1&2 are normal with a healthy status.
• Drive 3 is not in the storage pool.
• Drive 4 under allocation status says “System Partition Failed” with a health status.

I do see Drive 3 under HDD as not initialized.

Any advice on how I can restore by previous storage pool without loosing any data?
Thanks for any advice, guidance or recommendations.

 

[Rusty]

Welcome to the forum!

Well with your current situation you pulled 2 drives from a 4 drive SHR setup essentially breaking the 1-drive redundancy in the process.

Considering that the pool is degraded, whats the status of the data? If you can still access it, immediately start to back it up to spare disk(s). Once that is done, it would be best to redo the whole setup considering the damage. Also your 20TB incompatibility issue is also something that needs addressing, so be sure to have a compatible and detected drive before creating the new pool/volume.

 

[FAsnakes]

Thanks so much for the welcome and reply, makes sense and a painful lesson for me. I can still see the data although clicking around I’ve noticed a couple of files are corrupt. The volume is (read only).

is there any data I should be backing up on the two 12TB drives (the one that’s not initialized) and the other that Is showing system partition failed, or are they lost drives?

If lost, could a long shot be repairing drive 3 (which will erase it) and then trying to pull back in Drive 4 an an attempt to build back the SHR?

 

[Rusty]

Best to backup to a drive that has 0 relation to the current array. So connect a drive over USB, backup and then nuke the volume and start fresh. Finally, restore data back.

 

[Telos]

Here’s what I did:

1. Replaced drive 3 with a 20TB Ironwolf HD and it was not detected.
2. I then took the 20TB drive that wasn’t detected and put it in Drive 4.

If lost, could a long shot be repairing drive 3

You removed drive3, then subsequently drive4. So, your array lost 2 drives. If SHR or RAID 5, your entire pool is destroyed. As Rusty said, you must start over. You are beyond repair.

 

[FAsnakes]

Just circling back to close the loop. I contacted segate and synology support and they were pretty good. The Ironwolf 20TB drive was DOA, the replacement was delivered the next day and worked just fine. Synology was pretty optimistic that I should be able to recover. As soon as I borked everything the system went into read only mode (which I'm sure saved / preserved the data).

On Synology's advice I initialized and repaired Drive 3 and once that was complete I installed and repaired the new replacement 20TB in Drive 4 (this took around 48 hours to repair both drives). once that was done, I switched the drive to read / write (from read only). All of my installed packages in package center needed to be repaired which I did. Once I did that I was back up and running with the exception of a few packages that gave me this error: "This package requires you to enable "pgsql-adapter.service"

I followed the instructions here to resolve the issue.

Everything's up and running with no data loss. I'm doing a fresh data scrub, and will order another drive along with a USB enclosure so I can setup hyper backup along with B2 for cloud storage for my most important data. Thanks Rusty and Telos for the advice and help!

 

[Rusty]

Everything's up and running with no data loss

Well thats good news indeed! Glad both sides were so helpful on the matter, and that it was a defective drive not an incompatible one.

 

[Coop777]

Just circling back to close the loop. I contacted segate and synology support and they were pretty good. The Ironwolf 20TB drive was DOA, the replacement was delivered the next day and worked just fine. Synology was pretty optimistic that I should be able to recover. As soon as I borked everything the system went into read only mode (which I'm sure saved / preserved the data).

On Synology's advice I initialized and repaired Drive 3 and once that was complete I installed and repaired the new replacement 20TB in Drive 4 (this took around 48 hours to repair both drives). once that was done, I switched the drive to read / write (from read only). All of my installed packages in package center needed to be repaired which I did. Once I did that I was back up and running with the exception of a few packages that gave me this error: "This package requires you to enable "pgsql-adapter.service"

I followed the instructions here to resolve the issue.

Everything's up and running with no data loss. I'm doing a fresh data scrub, and will order another drive along with a USB enclosure so I can setup hyper backup along with B2 for cloud storage for my most important data. Thanks Rusty and Telos for the advice and help!

You got lucky, consider this a lesson learned. Always have an external backup of your data....

 

[FAsnakes]

You got lucky, consider this a lesson learned. Always have an external backup of your data....

I do, and i will.