Hello there,

 

first i'll try it in english  please ask if something could not be understood.

 

Okay here is what I have done:

 

I had an array with looking like this:

4 TB Parity

4 TB Datadisk

3 TB Datadisk

3 TB Datadisk

160 GB Datadisk

160 GB Datadisk

 

Of course the two 160 GB Disks are very small and as the array was only filled up to half of the last 3 TB Disk i decided to change the 160 GB Disks with greater ones.

I removed the two disks after taking the machine to power down. I installed two 5TB Disks.

 

After reboot there was the message array error and that 2 disks are missing. As i read into the wiki i go to tools and set "Tools -> New Config".

After that i assigned the disks like before in the correct places. And start the array.

 

Now the two 3 TB Disks are marked as unmountable and after the parity rebuild (10 h later) a whole bunch of data is missing.

 

What was wrong and what to do to repair everything?

 

The last thing i want (but i could) is to copy over network all my backuped data from another very slow cifs-machine....

 

It can be the "Auto" filesystem issue, stop the array, click those disks and change fs from auto to the correct fs, if it doesn't work post diagnostics.

Also list the disks as they are assigned now. You say you replaced the two 160's with 5TB, but since you only have 4TB parity that clearly can't be possible.

Because you added 2 x 5tb drives and your parity was only 4 tb i think that was your problem.

the parity drive must always be same size or larger than you largest array data disk.

As I understand it you should have removed the 4 tb parity first then replaced that with the 5tb drive.

Then resynced the parity to the new parity drive.

Then you should have removed 1 of the 160 gig drives and replaced it with  the other 5tb drive.

Then unraid should have rebuilt to replace missing data from the missing drive. (although empty)

then you should have removed second 160gig drive and replace it with the old 4tb parity

Then same as before unraid should have rebuilt to replace missing data from the missing drive. (although empty)

 

That is what went wrong is you removed 2 drives at once and didnt upgrade your parity to the larger disk size first??

 

I dont know how you can recover your lost data, but i am sure someone here will be able to tell you best way.

Also list the disks as they are assigned now. You say you replaced the two 160's with 5TB, but since you only have 4TB parity that clearly can't be possible.

 

I understood that since he couldn't start the array with the new disks he did a new config with the old ones, if that's not the case then yes, post a screenshot.

Also list the disks as they are assigned now. You say you replaced the two 160's with 5TB, but since you only have 4TB parity that clearly can't be possible.

 

I understood that since he couldn't start the array with the new disks he did a new config with the old ones, if that's not the case then yes, post a screenshot.

After a reread, you're correct, that's the way it looks anyway. Diagnostics zip and a screenshot would help clear things up.

It can be the "Auto" filesystem issue, stop the array, click those disks and change fs from auto to the correct fs, if it doesn't work post diagnostics.

That was my problem! Thank you all!

 

(I have put in the 5 TB disks but don't choose them for rebuilding, only i lay them into server )

 

After the rebuild is done i will take one of them for parity and follow up like gridrunner says.

 

Problem solved.

 

P.s. is it safe to use a wd green for parity? or better data?

It can be the "Auto" filesystem issue, stop the array, click those disks and change fs from auto to the correct fs, if it doesn't work post diagnostics.

That was my problem! Thank you all!

 

(I have put in the 5 TB disks but don't choose them for rebuilding, only i lay them into server )

 

After the rebuild is done i will take one of them for parity and follow up like gridrunner says.

 

Problem solved.

 

P.s. is it safe to use a wd green for parity? or better data?

Is there any data that you want to keep on the 2 160GB drives?

 

If not, it would be faster to once again set a new config, and assign all drives to the slots that you want. Then you can adjust the fs type for the drives that need it, and format the new drive and the old parity drive and rebuild parity based on the new layout instead of going through a multi step process to replace the 160's. Those steps are only needed if there is data that you want to keep on those drives.

 

The speed of the parity drive effects all write operations, but isn't used for read unless a drive fails. If you use your server for mostly read and few writes, parity speed doesn't matter much. If you write mostly, and multiple drives at once, then the parity speed is very important.

 

Safe is not a good description, either the drive is good or it is failing or failed. Drives should be tested before you trust them, then you can reasonably assume that they are safe to use.

 

All drives fail eventually, the trick is using them until just before they fail.

.... Drives should be tested before you trust them, then you can reasonably assume that they are safe to use.

 

All drives fail eventually, the trick is using them until just before they fail.

 

How to test for unraid? is there a plugin for?

.... Drives should be tested before you trust them, then you can reasonably assume that they are safe to use.

 

All drives fail eventually, the trick is using them until just before they fail.

 

How to test for unraid? is there a plugin for?

Preclear.

Unfortunately the documentation and instructions are a little confusing right now, since there are 2 different plugins with that name. The original is a front end for a command line script written ages ago, the beta is rewritten mostly from scratch.

 

You can also use the drive manufacturer's utility on a windows machine to run full diagnostics tests, it's not necessary to do testing in unraid. The preclear test is generally more thorough than the manufacturers test.

okay thank you for support.