Welcome to the Linux Foundation Forum!

What's 8e ? - Logical Volume Manager (LVM) Chapter

Hi, I cannot understand what's 8e in the next instruction:

"For example, assuming you have already created partitions /dev/sdb1 and /dev/sdc1 and given them type 8e"

What's "8e"? the next 8 commands to execute? a type of partition? I don't understand...

Another question is about the gvcreate command, What exactly the "physicalextensivesize" that we specified as param in this command:

sudo vgcreate -s 16M vg /dev/sdb1

It's necessary to specify this param for creating a group volume?



  • Hi Nicolas.

    Yes, in my understanding, 8e is a partition type of a Linux LVM, named "Linux LVM". Where for example, the standard Linux partition type is type 83, called "Linux". This is for actual physical drives. Check out $sudo fdisk -l and you will see something like this:

    Device     Boot   Start       End   Sectors   Size Id Type
    /dev/sda1  *       2048   1953791   1951744   953M 83 Linux
    /dev/sda2       1953792 629143551 627189760 299.1G 83 Linux

    Again, this is for the actual physical hard drives, before you start dealing with LVM.

  • coop
    coop Posts: 913

    to see the types possible you can run fdisk, then say "t" to change a type, it will ask a partition number and then you type "L" to see a list. Make sure you say q when done so you quit without a change. Note that in my experience the partition type seems to be ignored pretty much in Linux -- the operating system examines the partition to see what it is all the time anyway.

    I assume you meant vgcreate not gvcreate. If you don't specify the size of an extent with the -s parameter, you get get 4MB extents. The extent size is the chunk that the lvm is broken up into. A 1 GB file, for example, would that 256 extents.
    The bigger the extent size, the less fragmented the LVM would be but the less efficient it might be for moving things around and filling holes. I don' t know particularly good arguments for modifying the extent size from default but they probably exist. For SSD's fragmentation doesn't matter the way it does for physical hard drives.

  • Thanks, @coop . It looks like I was a little tired after some hours of study, but I'm back later, and I did understand.
    Thanks for your time and response.

  • Thanks@robbnl

  • rhin
    rhin Posts: 4

    if the device has a GPT partition table it's not '8e' for lvm and it's not 'fd' for raid, instead it's 31 and 29. This should be mentioned somewhere or changed, because imho MBR is in the process of dying.

  • Hi @rhin , what tool are you using? Can't find codes 29 and 31 on fdisk or gdisk. Please provide more details.


  • coop
    coop Posts: 913

    29 Linux RAID A19D880F-05FC-4D3B-A006-743F0F84911E
    30 Linux extended boot BC13C2FF-59E6-4262-A352-B275FD6F7172
    31 Linux LVM E6D6D379-F507-44C2-A23C-238F2A3DF9
    in fdisk on my laptop.

    I cannot research this right now, but on Linux the Type is ignored completely everywhere I have ever looked as the type is determined by analysis at run time by every utility including mount I often forget to even set the type as it is irrelevant. So in effect, this is curious but pedantic.

  • rhin
    rhin Posts: 4

    @luisviveropena said:
    Hi @rhin , what tool are you using? Can't find codes 29 and 31 on fdisk or gdisk. Please provide more details.


    It's just fdisk on my Ubuntu box.

    ❯ uname -a
    Linux blackbox 5.4.0-52-generic #57-Ubuntu SMP Thu Oct 15 10:57:00 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux

    but it's the same (29 & 31) on my Debian VPS:
    rh@elysium:~$ uname -a
    Linux elysium 4.19.0-12-amd64 #1 SMP Debian 4.19.152-1 (2020-10-18) x86_64 GNU/Linux


Upcoming Training