SOLVED disk identification

[dirkme]

This might sound funny, but if you have 5 disks in a raid, how do you identify the faulty drive (physically) in your NAS box?

 

[nojohnny101]

By serial number. You did note the location of your drives within your box by serial number correct?

You can get the serial number of each drive via the GUI (check manual). If the drive has dropped offline and is completely not visible in the GUI, then you can pull the drives (once the machine is shutdown) and find the serial number in the box that is not listed in the GUI.

 

[nightshade00013]

zpool status to see which disk is a problem.

Code:

zpool status
  pool: Zero-One
state: ONLINE
status: Some supported features are not enabled on the pool. The pool can
		still be used, but some features are unavailable.
action: Enable all features using 'zpool upgrade'. Once this is done,
		the pool may no longer be accessible by software that does not support
		the features. See zpool-features(7) for details.
  scan: scrub repaired 0 in 14h11m with 0 errors on Mon Jul 24 17:11:41 2017
config:

		NAME											STATE	 READ WRITE CKSUM
		Zero-One										ONLINE	   0	 0	 0
		  raidz3-0									  ONLINE	   0	 0	 0
			gptid/6d146e61-cedb-11e5-a60a-00259035884a  ONLINE	   0	 0	 0
			gptid/6e0d834b-cedb-11e5-a60a-00259035884a  ONLINE	   0	 0	 0
			gptid/6f07343c-cedb-11e5-a60a-00259035884a  ONLINE	   0	 0	 0
			gptid/700885e9-cedb-11e5-a60a-00259035884a  ONLINE	   0	 0	 0
			gptid/72778cc2-cedb-11e5-a60a-00259035884a  ONLINE	   0	 0	 0
			gptid/75416422-cedb-11e5-a60a-00259035884a  ONLINE	   0	 0	 0
			gptid/782f38a6-cedb-11e5-a60a-00259035884a  ONLINE	   0	 0	 0

glabel status to figure out which disk is connected to which interface/

Code:

glabel status
									  Name  Status  Components
gptid/6e0d834b-cedb-11e5-a60a-00259035884a	 N/A  ada0p2
gptid/6f07343c-cedb-11e5-a60a-00259035884a	 N/A  ada1p2
gptid/700885e9-cedb-11e5-a60a-00259035884a	 N/A  ada2p2
gptid/72778cc2-cedb-11e5-a60a-00259035884a	 N/A  da0p2
gptid/75416422-cedb-11e5-a60a-00259035884a	 N/A  da1p2
gptid/782f38a6-cedb-11e5-a60a-00259035884a	 N/A  da2p2
gptid/6d146e61-cedb-11e5-a60a-00259035884a	 N/A  da3p2
gptid/2693f5c7-27a4-11e7-aa52-00259035884a	 N/A  da4p1
gptid/47acfa41-f8c8-11e4-ad48-00259035884a	 N/A  da5p1
gptid/47bd8d36-f8c8-11e4-ad48-00259035884a	 N/A  da5p2
gptid/47ee65bb-f8c8-11e4-ad48-00259035884a	 N/A  da6p1
gptid/47fef3f8-f8c8-11e4-ad48-00259035884a	 N/A  da6p2

smartctl -a /dev/ada0 | grep ^Serial to find the serial number of the drive connected to a particular interface. The adaX or daX would be done for which ever drive is in question. you will then have the serial number which can be found on the drive.

Code:

 smartctl -a /dev/ada0 | grep ^Serial
Serial Number:	NK1334PEHLBR4S

 

[Stux]

Serial numbers are normally written on the front and top of each disk.

Btw, view disks in GUI will show the serial I believe

 

[dirkme]

Alright, all of you helped me a lot.

Just have to shutdown my FreeNAS to write down the serial numbers ;-)

Thank you so much guys.

Still don't know how to mark this topic solved.

 

[nightshade00013]

You can get the serial numbers from the Storage tab and then view disks as well unless you are wanting to label each disk as well for easier viewing.

To mark solved I believe you can edit your original post and change the heading.

 

[melloa]

Just have to shutdown my FreeNas to write down the serial numbers ;-)

And create a doc with physical location and serial #s :)

 

[nightshade00013]

Well this post inspired me to make a resource so if anyone later on comes across this looking for information I have two ways posted for you to figure it out. Via the GUI or Command Line.

https://forums.freenas.org/index.php?resources/identify-your-drives-by-serial-number.64/

 

[dirkme]

Well this post inspired me to make a resource so if anyone later on comes across this looking for information I have two ways posted for you to figure it out. Via the GUI or Command Line.

https://forums.freenas.org/index.php?resources/identify-your-drives-by-serial-number.64/

This is actually a brilliant guide for using the GUI, thanks a lot.

 

[nightshade00013]

This is actually a brilliant guide for using the GUI, thanks a lot.

Thank you I was hoping it would be easy to follow. I figure it's a lot easier to refer people to a guide than to explain it over and over again.

 

[Chris Moore]

This is actually a brilliant guide for using the GUI, thanks a lot.

One thing you might want to think about doing is labeling the drives.

Ex:

In a previous build, I could not see the serial numbers, so I put labels on the back corner of each drive with just the last four digits of the serial number, which is usually enough to positively identify a drive. The leading H or W was to differentiate between my HGST and WD drives in this particular build.

 

[Stux]

just the last four digits of the serial number, which is usually enough to positively identify a drive

Like that idea.

 

[daimon]

One good suggestion I read on another forum is you could write the bay number in the FreeNAS disk description. What else is the description field there for anyway?

However, what I am looking for is any kind of hardware support for disk (bay) identification. I'm sure many of the FreeNAS community members use some kind of hot-plug encasements. Mine for instance comes with two LED indicators for each bay - one indicates that the disk is connected, the other indicated data traffic. To be honest I don't have the slightest on what triggers these indicators, the encasement seems to have additional circuitry between power/sata cables and the disks, so I was hoping there was a bare minimum protocol that would let the "connection" indicator go blinking or something, based on the status detected by the OS. Back in the day when SAN and NAS were enterprise only and RAID was done exclusively by SCSI controllers this kind of a feature was a given. The last off-the-shelf encasement I bought (by PROMISE) still had this feature, kind of. It was a horrible product though, raid5 a joke, entire pools lost because of one failed disk, no diagnostics whatsoever and so on, but I digress. I don't want to shut down the NAS just because a disk failed.

So, in conclusion to my rambling, does FreeNAS have any kind of support for any kind of LED bay status indicators?

 

[Chris Moore]

To be honest I don't have the slightest on what triggers these indicators, the encasement seems to have additional circuitry between power/sata cables and the disks, so I was hoping there was a bare minimum protocol that would let the "connection" indicator go blinking or something, based on the status detected by the OS.

FreeNAS doesn't do this because it doesn't know what kind of hardware you have. For that matter, we really can't offer you any advice because we don't know what hardware you have either. In some drive chassis, there is a way to blink the lights by issuing a manual command also, you can find the bay information on some chassis by asking the controller. But your options depend on the hardware you are using, which is why the forum rules say that you are supposed to share your hardware info when you ask a question.
Please review the rules here: https://forums.freenas.org/index.php?threads/updated-forum-rules-4-11-17.45124/

 

[daimon]

FreeNAS doesn't do this because it doesn't know what kind of hardware you have.

Alrighty then, this answers my question. I suppose.

But your options depend on the hardware you are using, which is why the forum rules say that you are supposed to share your hardware info when you ask a question.
Please review the rules here: https://forums.freenas.org/index.php?threads/updated-forum-rules-4-11-17.45124/

Oh OK. I don't really have an issue so I didn't think HW specs were relevant. In any case, here:

Encasement: Eolize SVD NC11 4 Mini ITX PC Case (4x 3,5 HDD, 2x USB 2.0) for NAS System
MB: Gigabyte GA-J3455N-D3H Dual LAN Mini-ITX Motherboard with 16GB RAM
Disks: 3x WD Blue 3TB, 1x WD Red 3TB
Placed in an array with 1 redundant disk.
11.2 booting from a 16GB USB3.0 thumb. Home usage as a media server.

Unfortunately I got absolutely no documentation on the hot-plug circuitry with the box (bought on Amazon from DE). Nothing on the web either. So either some widely adopted protocols on what can be signaled over SATA exist and are hopefully supported by both the circuitry and the OS or else nothing. As you already explained, nothing it is.

Nevertheless, thanks for the info. Sincerely appreciated.

 

[Chris Moore]

I don't really have an issue so I didn't thin HW specs were relevant.

It may not be a problem, but I got the impression that you wanted information about if it were possible and how it might be accomplished.
The enclosure you have has a backplane that is simply a SATA pass-through. Some larger server chassis, such as the ones I manage at work, have SAS enclosure management services (SES) that can be used to blink the lights and control fan speed and such. That is not available on your system. You will need to track the drive serial number / location to know which drive has failed and be able to remove and replace the correct drive. Sorry.

 

[daimon]

Ok, again thanks for the info.

What do you think about the other idea, to write the bay number (I have only four so little room for confusion) into the disk description? That should remain associated with the disk and its serial number yes?

 

[Chris Moore]

Ok, again thanks for the info.

What do you think about the other idea, to write the bay number (I have only four so little room for confusion) into the disk description? That should remain associated with the disk and its serial number yes?

I write the last four digits of the serial number on the disk trays even on my rack server:

It is the most reliable way, because the description block in the GUI can become erased under certain circumstances.

I don't know why the photo is upside down.

 

[Stux]

This might sound funny, but if you have 5 disks in a raid, how do you identify the faulty drive (physically) in your NAS box?

If your drive bays have activity lights you can force activity on the drive you’re trying to identify, or if you can’t, say because it’s dead, then on all the ones you’re not trying to indentify.

I do this by using dd to read from the drives in question and write to dev null.