[Bug 1949252] Re: Ubuntu installed with wrong ZFS partition alignment for 4k HDD

Erich Eickmeyer  1949252 at bugs.launchpad.net
Sat Oct 30 14:08:37 UTC 2021


Thank you for taking the time to report this bug and helping to make
Ubuntu better. Please execute the following command only once, as it
will automatically gather debugging information, in a terminal:

apport-collect 1949252

When reporting bugs in the future please use apport by using 'ubuntu-
bug' and the name of the package affected. You can learn more about this
functionality at https://wiki.ubuntu.com/ReportingBugs.

** Changed in: ubiquity (Ubuntu)
       Status: New => Incomplete

-- 
You received this bug notification because you are a member of Ubuntu
Foundations Bugs, which is subscribed to ubiquity in Ubuntu.
https://bugs.launchpad.net/bugs/1949252

Title:
  Ubuntu installed with wrong ZFS partition alignment for 4k HDD

Status in ubiquity package in Ubuntu:
  Incomplete

Bug description:
  Using: Ubuntu 21.10 and OpenZFS 2.0. I did the standard ZFS install of Ubuntu.
  I had terrible bad response times from my brand new 2 TB Seagate STM2000M007 HDD, sometimes it takes 5 seconds or so to start the terminal and the system boots in 4 minutes from the HDD. I installed Ubuntu 21.04 on OpenZFS 2.0.2 and upgrade to Ubuntu 21.10 on OpenZFS 2.0.6 and both combinations on that HDD did cause that awful performance. It is a normal laptop drive at 5400 rpm. Its sequential read/write speeds are ~140MB/s while its seek times are ~13 msec.

  The measurement with gnome disk utility confirmed those values from
  the specification, the throughput on ZFS was ~120MB/s, but the
  difference could be caused by the latency of the 10 year old i5-2520M
  (2C4T; 2.5/3.2GHz). The disk SMART data was OK.

  Everything seems normal, ashift is set to 12 and the HDD has 4096 Byte
  physical sectors and 512 Byte logical sectors. All partitions are
  starting on 2048 boundaries of UEFI. I had been thinking about
  misalignment, because the throughput as measured by Conky was a factor
  2 larger than measured by ZFS itself (zpool iostat). The disk read-out
  is

  Disk /dev/sdd: 1.82 TiB, 2000398934016 bytes, 3907029168 sectors
  Disk model: 007-1R8174
  Units: sectors of 1 * 512 = 512 bytes
  Sector size (logical/physical): 512 bytes / 4096 bytes
  I/O size (minimum/optimal): 4096 bytes / 4096 bytes
  Disklabel type: gpt
  Disk identifier: F4581F88-2B2E-4C16-8D73-E55EC05CF987

  Device Start End Sectors Size Type
  /dev/sdd1 2048 4095 2048 1M BIOS boot
  /dev/sdd2 4096 1054719 1050624 513M EFI System
  /dev/sdd3 1054720 5249023 4194304 2G Linux swap
  /dev/sdd4 5249024 9443327 4194304 2G Solaris boot
  /dev/sdd5 9443328 3907029134 3897585807 1.8T Solaris root

  Many partitions start on the 2048 Byte boundary or a multiple thereof,
  which means that you have a misalignment and you double the number of
  IO Operations needed to read/write ZFS records. The last ROOT
  partitions starts at 4k sector number 2305½. In my opinion a serious
  bug in the ZFS implementation of Ubuntu, however most people will not
  have the issue, because they run mostly from 512 Byte aligned SSDs or
  nvme-SSDs.

  I think it should be a general rule on 4k HDDs and SSHDs you should
  align partitions on 4K boundaries.

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1949252/+subscriptions




More information about the foundations-bugs mailing list