drive error - replacing a drive when the machine has no more physical space?

[mrklaw]

Hi,

 

Getting a lot of read errors on my disk 2. First time since I built this server which was a few years ago now. So I figure I might as well replace the drive and take the opportunity to upgrade the space from 2-4TB. Then consider looking to upgrade the other drives over the next 12-18 months.

 

But I have no more physical space to attach new drives, so does that mean I can't preclear them - my only real choice is remove the faulty drive, put the new one in and just have unraid sort it out (with the array offline)? Any way around that? I have a crappy old laptop that I don't use anymore in case I could use that to boot a version of unraid just for preclearing, perhaps with a USB-Sata enclosure/dock?

 

Any suggestions gratefully received.

[trurl]

Hi,

 

Getting a lot of read errors on my disk 2. First time since I built this server which was a few years ago now. So I figure I might as well replace the drive and take the opportunity to upgrade the space from 2-4TB. Then consider looking to upgrade the other drives over the next 12-18 months.

 

But I have no more physical space to attach new drives, so does that mean I can't preclear them - my only real choice is remove the faulty drive, put the new one in and just have unraid sort it out (with the array offline)? Any way around that? I have a crappy old laptop that I don't use anymore in case I could use that to boot a version of unraid just for preclearing, perhaps with a USB-Sata enclosure/dock?

 

Any suggestions gratefully received.

A rebuild does not require a clear disk since it is just going to be overwritten anyway. The only time a disk is required to be clear is when you are adding it to a new slot in a parity protected array. This is so parity will remain valid. A clear disk is all zeros so has no impact on parity. Many people preclear new disks even when it is not required just to test them. Often the drive manufacturer will also have software for testing disks if you want to do it that way.

 

Do you already have a 4TB parity disk? Parity must be at least as large as the largest data disk.

 

Also, errors may not really indicate something wrong with the drive. Very often it is something else like connections or cables (SATA and power) to the drive. See v5 help in my sig for how to get a SMART report for the drive.

[mrklaw]

Appreciate the simple language, thanks

 

All my drives are 2TB, so does that mean I would need to upgrade my parity drive at the same time? I just figure it would be a waste to replace with another 2TB drive, and I was looking to slowly upgrade the space anyway so I figured it was a good time to start.

 

I got a smart report, which I've attached. Summary of first lines is

ID# ATTRIBUTE_NAME          FLAG    VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE

  1 Raw_Read_Error_Rate    0x002f  178  178  051    Pre-fail  Always      -      45139

  3 Spin_Up_Time            0x0027  170  168  021    Pre-fail  Always      -      6483

  4 Start_Stop_Count        0x0032  098  098  000    Old_age  Always      -      2112

  5 Reallocated_Sector_Ct  0x0033  200  200  140    Pre-fail  Always      -      0

 

I was noticing it stuttering during playback of a video. Managed to eventually copy that file off to a local drive on my PC but it took a couple of retries.

 

edit: if the disk does need replacing, am I stuck replacing it with a 2TB? I don't think I should update the parity drive to a larger drive until my disk2 is sorted. Unless I can use a 4TB drive but only 2TB will be used - then update parity to 4TB, then redo disk 2 to a full 4TB?

 

smart.txt

[itimpi]

You cannot simply upgrade the parity drive to a larger one while there I a problem with a data disk.  Having said that there is a process called "Swap-Disable" that allows you to simultaneously upgrade the parity drive to a larger size, and use the old parity drive to rebuild the failed/problem disk. 

 

If you do not want to do that, then you can only replace the failed drive with a 2TB one.  You might want to include the rest of the SMART report to see if it is showing issues in case you just have something like a cabling problem as there is no point in buying a new drive if you do not need it..  Also supplying a syslog would be a good idea to help with diagnosis.

[mrklaw]

I've ordered a new 2TB drive for now - prices are cheap enough to do that and then look at upgrading later.

 

I attached the smart report to my previous post - is that not the correct one? I'll grab a syslog in a short while - is that just copy/paste everything from the syslog page into a text file and attach?

 

edit: syslog attached.

syslog-2016-01-22.zip

[JorgeB]

197 Current_Pending_Sector  0x0032   196   196   000    Old_age   Always       -       1458

 

Disk has bad sectors and should be replaced asap.

 

Like itimpi pointed out, you can use the parity swap procedure.

 

https://lime-technology.com/wiki/index.php/The_parity_swap_procedure

 

[mrklaw]

thanks. Think I'll keep it simple® and just swap the failing drive.

 

Are there any tools that can regularly do a smart report and flag potential failures, or is it just something I should do regularly myself

[JorgeB]

Update to v6, system notifications will alert you to most SMART issues.

[mrklaw]

Update to v6, system notifications will alert you to most SMART issues.

 

Fantastic. Will upgrade to v6 just as soon as I've switched my drive out.