Bellow is the current status of my new raid array.  I have a share named `storage` and i'm currently rsyncing data from an older FreeNAS machine into the `/mnt/user/storage` to copy all of the data to the new Unraid machine.  I'm hoping that that's an ok way to transfer the data from another machine to Unraid - i assumed that `/mnt/user` mount was where the array was mounted.  But upon looking at the image below you can see that most (all?) of the data is on the first disk.  Is this normal?  Will it get spread around later?  Is this not how I should be transfering data?

 

5 minutes ago, Brian Dilley said:

Will it get spread around later?

Depends on the allocation mode and split level set for the share(s), default is high water and split always.

Here's the settings:

 

 

12 minutes ago, Brian Dilley said:

new raid array

Unraid IS NOT RAID.

 

Each disk is an independent filesystem. Each disk can be read independently on any linux. Each file exists completely on a single disk. Folders can span disks, this is called User Shares.

 

Reads from the parity array is at the speed of the single disk which contains the file. Writes to the parity array are somewhat slower due to realtime parity updates.

 

Unraid is not as fast as RAID, but it has other benefits. Since each disk is an independent filesystem

• if you lose more than parity can recover, you haven't lost everything because all good disks still have their complete data
• you can use different sized disks in the parity array
• you can easily add more disks to the array without rebuilding the whole array
• you can easily replace disks with larger disks

As mentioned, highwater allocation is the default, and for good reason. It is a compromise between using all disks eventually without constantly switching between disks just because one disk temporarily has more free space.

 

Unraid does not automatically move files between array disks, so files already written do not get spread around. When allocation method says disk1 is done, it will begin writing new files to another.

 

Since all your disks are 12TB highwater is easy to figure out, it is half that or 6TB. When 6TB have been written to disk1, the next disk in line will be chosen until it has reached highwater, etc.

I read the definition of highwater - so it seems to be working correctly.  Is this safer/better/faster/etc. than the "most-free" option?

Just now, Brian Dilley said:

I read the definition of highwater - so it seems to be working correctly.  Is this safer/better/faster/etc. than the "most-free" option?

Just now, trurl said:

highwater allocation is the default, and for good reason. It is a compromise between using all disks eventually without constantly switching between disks just because one disk temporarily has more free space.

 

 

Thanks!

4 minutes ago, Brian Dilley said:

than the "most-free" option?

Most free is bad for performance since writes to parity disk will overlap when changing data disk, resulting in worse write performance.