fachinformatiker-wiki

it's easy when it's here

User Tools

Site Tools


linux:filesystem:zfs

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision
Previous revision
linux:filesystem:zfs [2025/10/08 01:15] gsyslinux:filesystem:zfs [2025/10/08 03:44] (current) – [Format NVMe with bigger sector size] gsys
Line 45: Line 45:
 </code> </code>
  
 +by-id
 +<code>
 +zpool create -o ashift=12 -O compression=lz4 -O atime=off -O xattr=sa local-datastore-nvme01 \ 
 +mirror /dev/disk/by-id/nvme-SAMSUNG_MZQL23T8HCLS-00A07_S3R14LNUMB3R1 /dev/disk/by-id/nvme-SAMSUNG_MZQL23T8HCLS-00A07_S3R14LNUMB3R2 \
 +mirror /dev/disk/by-id/nvme-SAMSUNG_MZQL23T8HCLS-00A07_S3R14LNUMB3R3 /dev/disk/by-id/nvme-SAMSUNG_MZQL23T8HCLS-00A07_S3R14LNUMB3R4
 +</code>
 ===== Delete ZFS Pool ===== ===== Delete ZFS Pool =====
  
Line 63: Line 69:
 Check column "Fmt". "+" marks the current format. "Rel_Perf" indicates the higher performance low level formats. The lower the better. 0 being the best. Check column "Fmt". "+" marks the current format. "Rel_Perf" indicates the higher performance low level formats. The lower the better. 0 being the best.
 <code> <code>
-root@pve01:smartctl -a /dev/nvme4n1+root@pve01: smartctl -a /dev/nvme4n1
 {...} {...}
 Supported LBA Sizes (NSID 0x1) Supported LBA Sizes (NSID 0x1)
Line 79: Line 85:
 Verify the change. Verify the change.
 <code> <code>
-root@pve01:smartctl -a /dev/nvme4n1+root@pve01: smartctl -a /dev/nvme4n1
 {...} {...}
 Supported LBA Sizes (NSID 0x1) Supported LBA Sizes (NSID 0x1)
Line 88: Line 94:
 </code> </code>
 https://openzfs.github.io/openzfs-docs/Performance%20and%20Tuning/Hardware.html#nvme-low-level-formatting https://openzfs.github.io/openzfs-docs/Performance%20and%20Tuning/Hardware.html#nvme-low-level-formatting
 +
 +<code>
 +lsblk -d -o NAME,PHY-SeC,LOG-SEC
 +nvme id-ns -H /dev/nvme4n1 | grep 'LBA'
 +</code>
 +
 +===== Misc =====
 +
 +==== Remove orphaned pool ====
 +
 +Debian fails to start import service of an orphaned pool.
 +<code>
 +zpool set cachefile=/etc/zfs/zpool.cache poolname    # set cache file for running pools
 +update-initramfs -k all -u
 +reboot
 +</code>
 +
 +check labels on disks
 +<code>
 +zdb -l /dev/nvme1n1p1
 +</code>
linux/filesystem/zfs.1759878953.txt.gz · Last modified: by gsys