How to decode disk serial number

[Alecmascot]

zpool clear poolname

 

[jgreco]

But it is interesting that a critical alarms remains even though the disk were plugged back in

No, that's an alert. An alert is a message telling you that something happened. It does not necessarily mean that the situation is still going on.

If you had a disk that had a problem running a SMART test last week, but wasn't ACTUALLY having a problem doing that operation RIGHT THIS VERY SECOND because the test had failed and was now "done", you would nevertheless want to receive an alert that it had failed, ... right?

You are expected to acknowledge the issue the NAS is trying to let you know about, and then it is dismissed. You are expected to understand the implications; the software isn't going to try to hide potential problems from you, and it isn't really a good idea to have the software try to automatically triage potentially serious problems, because it is more likely to be wrong than right. You are the only one in the loop who can say "oh yea right that's because I disconnected some drives, no real problem here."

 

[john60]

zpool clear poolname

I assume you are saying that this will get rid of the red unhealthy x.

The man page says

A pool has at least one device with ZFS metadata cleared by using zpool label -C command. A cleared pool cannot be imported.

What does cannot be imported mean? Should I be concern before trying this command?

 

[Alecmascot]

zpool label -C is something else ....

zpool clear [-F [-n]] pool [device]

Clears device errors in a pool. If no arguments are specified, all device errors within the pool are cleared. If one or more devices is
specified, only those errors associated with the specified device or devices are cleared.

-F Initiates recovery mode for an unopenable pool. Attempts to discard the last few transactions in the pool to return it to an
openable state. Not all damaged pools can be recovered by using this option. If successful, the data from the discarded transac-
tions is irretrievably lost.

-n Used in combination with the -F flag. Check whether discarding transactions would make the pool openable, but do not actually
discard any transactions.

 

[john60]

No, that's an alert. An alert is a message telling you that something happened. It does not necessarily mean that the situation is still going on.

If you had a disk that had a problem running a SMART test last week, but wasn't ACTUALLY having a problem doing that operation RIGHT THIS VERY SECOND because the test had failed and was now "done", you would nevertheless want to receive an alert that it had failed, ... right?

You are expected to acknowledge the issue the NAS is trying to let you know about, and then it is dismissed. You are expected to understand the implications; the software isn't going to try to hide potential problems from you, and it isn't really a good idea to have the software try to automatically triage potentially serious problems, because it is more likely to be wrong than right. You are the only one in the loop who can say "oh yea right that's because I disconnected some drives, no real problem here."

Good point, an alert can be cleared by the user when he/she corrects something.
Thank you for pointing this out. That makes a lot of sense.

zpool label -C is something else ....

yikes, thanks so much for pointing that out. I must be more careful.
So "zpool clear poolname" in the console will clear the GUI unhealthy red x?
Is there an equivalent command in the web GUI?


Still open questions from my experimenting/fire drill:
- funky serial numbers on the first experiment ... need to try again with the same 2 disks

 

[john60]

I pulled out the same drives as I first did. The serial numbers in the alert don't match the serial number of the removed drives ( see the first images I posted) on this thread.

The serial number in the "Storage / Disks" does show the still attached drives.

So in the case of complete disk failure, I can simply look for drives missing from this list.

 

[jgreco]

The serial numbers in the alert

There are no serial numbers in the alert. It's a ZFS internal identifier.

If you want to argue that the alert could be improved, then we agree. ;-)

 

[Jessep]

Well this is actually a very good question. I'm reasonably positive the GUI used to indicate this somewhere. I just checked one of our filers known for "the long scrubs" and while it is clearly doing a scrub as indicated from the CLI, I am not seeing an indication of this in the GUI. Perhaps I am just not yet fully awake this morning...?

Scrubs show up in the same place.

 

[joeschmuck]

What would be a nice future feature, custom GUI screens where you can select from a list of items to display on a custom tab or maybe 2 or 3 tabs. This would make it nice for people to personalize the GUI for the items they just want to see vice mucking around all the menus. It took me some time to locate the SMART Test configuration screen in the Scale version. It was not obvious to me. But that is not what this thread is about, just a thought.

 

[john60]

There are no serial numbers in the alert. It's a ZFS internal identifier.

If you want to argue that the alert could be improved, then we agree. ;-)

Not arguing anything, just trying to understand how it works before I go live with real data/users.

The observation is not that funky number is reported, but the funky number is reported sometimes, and real serial numbers
other times.
When I pull the top 2 disks, I get this ZFS # that does not match with anything I could find about a disk.
When I pull the bottom 2 disks, I get serial numbers in this same alert that do match disk serial numbers.

Not a big deal with a 6 disk NAS, but could be a more serious issue with a large number of disks.
Does Truenas development team want tickets for improvement like these especially since I can reproduce this anomaly?

A more interesting thing is that the red unhealthy does not go away with a pool scrub and I can find no clear button in the GUI.

but a restart of the NAS will clear this out-of-date information.

Taking down the NAS to update / remove an out-of-date GUI indication does feel more serious.
Think I should report this finding?

 

[jgreco]

funky number is reported sometimes, and real serial numbers
other times.

Then don't look around too much, because you can also find references to device names in some places, and GUID's in even others. ;-) Consistency is nice, but it isn't always possible for reasons that aren't always clear to end users.

When I pull the top 2 disks, I get this ZFS # that does not match with anything I could find about a disk.
When I pull the bottom 2 disks, I get serial numbers in this same alert that do match disk serial numbers.

Not a big deal with a 6 disk NAS, but could be a more serious issue with a large number of disks.
Does Truenas development team want tickets for improvement like these especially since I can reproduce this anomaly?

It doesn't hurt to report it. They will likely want a debug dump if they are interested in the issue, so I recommend including one.