Conversation

Notices

1. Embed this notice
   Polychrome :blabcat: (polychrome@poly.cybre.city)'s status on Tuesday, 05-Mar-2024 04:48:45 JST Polychrome :blabcat:
   Trying to help a person with PopOS and an Intel SSD and for some reason Gnome Disks is going wild over the Program Fail Count Chip counter running well past '60135377235'.
   
   A quick search shows that Intel uses this SMART value to report PLP capacitor tests which is probably not what Gnome Disks is expecting, and this is causing PopOS to consistently warn that the disk is about to fail despite being brand new.
   
   Does anyone know how to get PopOS to work with PLP enabled Intel SSDs without going crazy?
   
   #linux
   • Embed this notice
     Polychrome :blabcat: (polychrome@poly.cybre.city)'s status on Tuesday, 05-Mar-2024 06:08:30 JST Polychrome :blabcat:

     in reply to

     • argv minus one
     @argv_minus_one probably the only way out of this, yeah.
     
     Unfortunately the one facing the error is a casual user with minimal understanding in computers so guiding them to generate the report will take some effort, but I'll give it a shot. :meowShrug:
   • Embed this notice
     argv minus one (argv_minus_one@mstdn.party)'s status on Tuesday, 05-Mar-2024 06:08:31 JST argv minus one

     in reply to

     @Polychrome

     So, software that tries to predict drive failure will instead look at the *raw* values and make its own conclusions…which, as you can see, are highly vendor-dependent and therefore highly likely to be very wrong.

     Anyway, this is a bug in whichever software is raising the false alert, presumably Gnome Disks. You should probably file a report, and include the output of `smartctl -x <device>`, so they can fix it.

     2/2

   • Embed this notice
     argv minus one (argv_minus_one@mstdn.party)'s status on Tuesday, 05-Mar-2024 06:08:32 JST argv minus one

     in reply to

     @Polychrome

     I was wondering if something like this was going to happen at some point.

     SMART attributes all have a normalized value, and a threshold beyond which the drive is declared to be failing. Drive makers have been ridiculously over-optimistic about setting these thresholds. For instance, if an HDD has *any* reallocated sectors, it's going to die soon, but it won't actually say it's dying until it's reallocated a ton of sectors and is basically already dead.

     1/2

   • Embed this notice
     argv minus one (argv_minus_one@mstdn.party)'s status on Tuesday, 05-Mar-2024 06:16:28 JST argv minus one

     in reply to

     @Polychrome

     I take it you don't have remote shell access to this machine? That's unpleasant. It's pretty hard to be somebody's IT administrator when you don't have administrative access to the IT.

   • Embed this notice
     Polychrome :blabcat: (polychrome@poly.cybre.city)'s status on Tuesday, 05-Mar-2024 06:16:28 JST Polychrome :blabcat:

     in reply to

     • argv minus one
     @argv_minus_one I might have to work something out at this rate :p