How to do a system reset of the SSD caching

[dstel]

Hello,
Because of the 5401 bug (Ticket 37294), I would like to know if there is a way, including by ssh commands, to reset the state of the drives so I can use them again.

I already rebooted, and I don't want to open the NAS to go format those drives on another server.

EDIT, SOLUTION (workaround)

Disclaimer: this is my workaround. I found it by doing some stuff on my NAS. It worked for me, it is not an official solution given by the support.

Connect with ssh.

Find in /dev/mapper maps that are not complete for the cache : you need:
mdX_m
mdX_s
and if successful you also have a
mdX_c
but if you have a problem, you don't see it.

The problem is, when it fails you only have a mdX_m and mdX_s but not the mdX_c and I guess the system refuse to recreate them when you do the creation with the menu.

So you have to :

Code: Select all

dmsetup remove mdX_m
dmsetup remove mdX_s

and reboot. Then you can use the menu to create a cache.

[orion]

Do you mean to remove SSD cache? Found a guide here: https://www.asustor.com/en-gb/online/College_topic?topic=202

(B) Do not remove SSDs without ejecting safely. If SSD Caching is no longer required, the SSD must be ejected first before removal. Click SSD Cache under Management. Do not turn off the NAS before the ejection completes.

[dstel]

No, it's something deeper because of a bug. I needed some ssh trick. But now after creation/deletion from the interface and weird period when I've got 0byte/0available on my volume and then a reset, I was able to create the SSD cache again. However there is still a potential issue for future user, so the ticket is still open. It's not an incident anymore, it's just a problem that need a fix.

[Johnson23]

Are you not afraid to destroy all data on the SSD? Be cautious!

[dstel]

Edited with the workaround.

[TheDesignated1]

Is the 5401 bug regarding SSD caching still a thing? I think I'm a "victim" of it.

Created SSD cache to attach to Volume 1. It creates, mounts, gives 5401 error, then lists my volume as 0B/0B.

Not sure how the workaround is executed (I'm still a noob in ssh/linux). I go to the /dev/mapper, but the only thing listed is "control."

Any help would be great. Thank you.