I'm preparing my "decommissioned" stationary PC to be used as an Unraid server.  It has a 256GB SSD boot drive and 4x1.5TB HDD data drives.  My question is what I should do with the data drives before installing Unraid?  I have a separate RAID controller that the drives are connected to, and the controller is connected so that it's just one big drive from the OS' point of view (which presently is Windows Server).  But I gather that this is not a good way to do it when using Unraid.  Should I rather split it up as four separate drives that are accessible to Unraid, so that Unraid then can distribute data across the drives as it sees fit?  How then with hardware support?  Will Unraid use the hardware support in the RAID controller to distribute data, or will it handle it by itself in software?

Unraid IS NOT RAID. Best if you don't even attempt to use a RAID controller with Unraid. Which controller is it? Can it be flashed to IT mode?

1 hour ago, ohj said:

But I gather that this is not a good way to do it when using Unraid.  Should I rather split it up as four separate drives that are accessible to Unraid, so that Unraid then can distribute data across the drives as it sees fit? 

Data drives in unRAID are separate disks each with its own independent file system.   No RAID.

 

User share configurations determine how files are allocated across the independent disks.  Shares can be limited to one disk, span all disks, or anything in between.  Allow the allocation method and split levels help determine how full disk get before spilling over to other disks and at what point content folders should be split and allowed to spill over to other disks the share is configured to use.

On 12/31/2020 at 12:25 AM, trurl said:

Which controller is it? Can it be flashed to IT mode?

Thank you for the response.  I can't remember, as I haven't touched the controller in years.  But I think I will check if I can split the RAID array up then, and just use it as a simple SATA interface with separate drives.

On 12/31/2020 at 1:43 AM, Hoopster said:

Data drives in unRAID are separate disks each with its own independent file system.   No RAID.

Thank you for the response.  I will split the RAID array up then and give it a try. 

Some people seem to resort to setting up each disk as a separate RAID. Don't do it that way. 

On 12/30/2020 at 6:25 PM, trurl said:

Which controller is it?

 

15 hours ago, trurl said:

Some people seem to resort to setting up each disk as a separate RAID. Don't do it that way. 

Thank you for the tip!  Out of curiosity, why not?  Additional overhead and increased latency?

RAID controllers don't always pass correct disk identifier (serial number) which can cause problems keeping disk assignments straight, and also don't pass SMART information, making it impossible for Unraid to monitor disk health and alert you when there are problems.

Yes, I noticed that.  I split up the RAID, and the separate drives were detected in Disk Management in Windows.  But I still weren't able to do anything with them.  But I had enough available SATA connectors on the mainboard, so it was fine.

 

I very much like what I see with unRAID so far.  I'm setting the drive array up with 5 x 2TB drives (not 1.5TB as I stated earlier) - 4 data drives and 1 parity drive.  But now when unRAID is preparing the drive array, I see one of the drives is showing some signs of having problems.  The error count currently shows around 24,000.  What would be the smart thing to do about this?  Should I remove the drive immediately, or will the bad sectors be flagged so the drive is safe to use?

 

Also, I see it takes ages and ages with the Parity-Sync / Data-Rebuild process.  It's been chewing at it for over four hours, and it says it has several days(!) left.  This is a new array and the drives have been formatted.  Why does this step take so long?  Is something wrong here (e.g. if the troubled drive mentioned earlier cause problems)?

 

Edited January 3, 20215 yr by ohj

I read up a little on the SMART tests, and ran SMART short tests on all the drives.  Two of them came up with errors during the tests, so I excluded them from the drive array, and restarted the parity build.  Now it's steady on ~150 MB/sec and should be finished in about three hours.

 

In regards to the drives, they are all fairly old drives that I bought for a Synology NAS perhaps 8 or 10 years ago.  So time to buy new drives, I'd say...

 

 

Since some of these drives failed even the SMART short rests I would strongly recommend yhat you run the extended tests on the remaining ones.   The short test is by no means definitive.

4 hours ago, ohj said:

I read up a little on the SMART tests, and ran SMART short tests on all the drives.  Two of them came up with errors during the tests, so I excluded them from the drive array, and restarted the parity build.  Now it's steady on ~150 MB/sec and should be finished in about three hours.

 

In regards to the drives, they are all fairly old drives that I bought for a Synology NAS perhaps 8 or 10 years ago.  So time to buy new drives, I'd say...

If you are running the latest Unraid 6.9.0 rc release then the Parity Check Tuning plugin can now restart array operations after a shutdown/reboot or array stop/start.

12 hours ago, ohj said:

In regards to the drives, they are all fairly old drives that I bought for a Synology NAS perhaps 8 or 10 years ago.  So time to buy new drives, I'd say...

Probably. The way Unraid uses parity protection, your data is only as safe as your weakest drive. All the sectors of all the drives are needed when reconstructing a failed drive, so having a known bad drive included in the array even though it's still "working" at the moment is a bad idea. Imagine this scenario... you purchase a pair of brand new drives, one for parity, one for data1. You decide that your gaggle of old drives are good enough, they haven't completely died yet, and what are the chances of two dying at the same time, right? Parity will allow you to rebuild to a new drive if one of the old ones fail, so you feel safe.

 

Until one of your brand new drives decided to fail, and that one weak old drive decides it can't handle all that stress of rebuilding, so you just lost both drives worth of data.

 

The parity check is a good tool to keep up with the health of your drives, if something feels off, like it did with that failing drive in place, you need to figure out why. If a parity check won't complete error free in a timely fashion, a rebuild of a failed drive won't either.

 

Also... not all drive issues are really disk failures, MANY times it's connections or power.

And in any case, parity is not a substitute for a backup plan. You must always have another copy of everything important and irreplaceable. Parity can help recover from a failed disk, but there a plenty of more common ways to lose data, including user error.

On 1/3/2021 at 9:10 PM, jonathanm said:

Probably. The way Unraid uses parity protection, your data is only as safe as your weakest drive. All the sectors of all the drives are needed when reconstructing a failed drive, so having a known bad drive included in the array even though it's still "working" at the moment is a bad idea. Imagine this scenario... you purchase a pair of brand new drives, one for parity, one for data1. You decide that your gaggle of old drives are good enough, they haven't completely died yet, and what are the chances of two dying at the same time, right? Parity will allow you to rebuild to a new drive if one of the old ones fail, so you feel safe.

Thank you so much for this!  I actually ended up with canning the whole set of drives.  They were almost 10 years old, so it was time to retire them anyway, I think.  Ended up with buying three new 4TB drives, having one of them as parity.

 

I notice you have two unRAID servers, one for ordinary use and one for backup.  Is the backup server an offsite server that you're performing backup to of the main server on a set schedule?  Would you recommend that rather than performing backup to a cloud service?

On 1/3/2021 at 10:44 PM, trurl said:

And in any case, parity is not a substitute for a backup plan. You must always have another copy of everything important and irreplaceable. Parity can help recover from a failed disk, but there a plenty of more common ways to lose data, including user error.

Very true!

6 hours ago, ohj said:

 

I notice you have two unRAID servers, one for ordinary use and one for backup.  Is the backup server an offsite server that you're performing backup to of the main server on a set schedule?  Would you recommend that rather than performing backup to a cloud service?

Yes, I have another server at a different address, it's not only for backup, but yes, the servers complement each other with the most important data. I don't bother keeping everything backed up, just the irreplaceable data.

 

Renting space on somebody elses computer (cloud backup) makes sense if you only have a small amount to back up, once you get beyond several TB it gets messy or expensive. I prefer to keep my data myself. Connection speed is also a primary concern for some.