Linus RAID failure: suspense / horror thriller for IT professionals (video)



  • Hah, good spot. typo of course, and a funny one 😄


  • Notification Spam Recipient

    @Zadkiel said:

    Hah, good spot. typo of course, and a funny one 😄

    Hey, if an IT department actually knows it is shit, that's like, three-fourths of the battle right there!


  • Discourse touched me in a no-no place

    @Zadkiel said:

    with an IT department that actually knows it's shit

    I'm fairly certain that stray apostrophe negates what you're trying to get across there...


  • Discourse touched me in a no-no place

    @Tsaukpaetra said:

    They wanted to have us put our copy of the source control tree on a network drive, and I could only shrug in confusion.

    The way to respond to that is to say that you're blocked from doing that until they upgrade the connection between the network storage and the machines that are using it to minimum gigabit bandwidth. Since they've not done that yet, you're blocked from doing your end. Put a formal request in and all; it's then up to them to sort things out to make it possible and you can get on with actually doing useful stuff.

    Either that or say that there are (theoretical) problems with getting locking working on networked filesystems. Which there most certainly are; network failures make the whole area a total custerfluck.



  • @Zadkiel said:

    I took one of the failed disks which was exhibiting signs of bearing failure in the motors, connected it to a 3m SCSI cable, plugged one end into the SCSI backplane where the drive had been, wrapped the other end in plastic, dragged a fridge/freezer from the company kitchen into the server room, drilled a hole through the side of the freezer and ran the SCSI cable through the hole so the wrapped drive was inside the freezer. This shrunk the bearings in the motor enough that it ran long enough to rebuild the other bad disk, and from there I could rebuild the 'freezer' disk and recover the whole array.

    That's a pretty amazing story.

    @Zadkiel said:

    His backup strategy. One single backup routine which operates by deleting the old backup and then writing a new one. And that's it. No offsite, no media rotation.

    This is the strangest bit.

    I mean, everything else is "whatever, that's just like your opinion, man".

    But flying without ANY backups? Amazingly irresponsible.



  • @PJH said:

    @Zadkiel said:
    with an IT department that actually knows it's shit

    I'm fairly certain that stray apostrophe negates what you're trying to get across there...

    Sigh. Fine. I'll edit it. I thought after the first person pointed it out and I confirmed it was a typo that there wasn't a need, but 2 more people pointing out the same spelling mistake proves me wrong.




Log in to reply