Skip to content

Latest commit

 

History

History
43 lines (36 loc) · 2.45 KB

3_SIMULATING_HARDWARE_FAILURE.md

File metadata and controls

43 lines (36 loc) · 2.45 KB
Raw

Simulating Hardware Failure Exercise Answers

  • Create an unmirroed Zpool called zfspool, remove /disks/disk0, catch the error in ZFS, confirm that the pool is utterly broken and that your files are unrecoverable.

    • Create the zpool:
      • zpool create zfspool /disks/disk0 /disks/disk1
      • populate-zfs-filesystem /zfspool/ 5 5
      • sha1-check-files /zfspool/ - Initial check
      • zpool status - Verify state is ONLINE
    • Break the disk:
      • Simulate breaking the disk with the command break-disk disk0
      • zpool scrub zfspool - Watch this hang so hard that even ctrl-C can't break out of it.
      • zpool status - Run this in another SSH session, watch it too hang and be immune to ctrl-C.
      • Run vagrant halt -f && vagrant up in another shell to kill and restart the VM, then vagrant ssh back in.
      • Run zpool status and see that no pools are available.
      • Enjoy this valuable learning experience about just how quickly you can lose all of your files on a Zpool with no redundancy.
    • Run zfs-add-disk-file disk0 1024 to re-create the corrupted disk for the next exercise.

  • Created a mirrored Zpool called zfspool, remove /disks/disk0, catch the error in ZFS, fix the pool, verify that files are unharmed.

    • Create the zpool:
      • zpool create zfspool mirror /disks/disk0 /disks/disk1
      • populate-zfs-filesystem /zfspool/ 5 5
      • sha1-check-files /zfspool/ - Initial check
      • zpool status - Verify state is ONLINE
    • Break the disk:
      • Simulate breaking the disk with the command break-disk disk0
      • zpool scrub zfspool - This completes quickly, and will catch the error
      • zpool status - disk0 now shows as DEGRADED
      • sha1-check-files /zfspool/ - Confirm files aren't corrupted
    • Replace the disk (option 1, add in a new disk):
      • zpool replace zfspool /disks/disk0 /disks/disk2
      • zpool status - disk0 has been removed from the pool, disk2 is now online
      • sha1-check-files /zfspool/ - Verify that the files are unharmed
    • Replace the disk (option 2, replace the existing disk):
      • zfs-add-disk-file disk0 1024 - Simulate replacing the physical disk
      • zpool replace zfspool /disks/disk0
      • zpool status - disk0 is back!
      • sha1-check-files /zfspool/ - Verify that the files are unharmed
    • Run zfs-add-disk-file disk0 1024 to (re)create the disk0 file when done with this exercise