Uploaded image for project: 'Lustre'
  1. Lustre
  2. LU-18514

confusing RAID information from mke2fs

Details

    • Question/Request
    • Resolution: Unresolved
    • Minor
    • None
    • Lustre 2.16.1
    • None
    • Alma Linux 9.5
    • 3
    • 9223372036854775807

    Description

      I am trying to create an ldiskfs partition, but the output of mke2fs is confusing:

       

      mkfs.lustre  --ost --mkfsoptions=" -E stride=16,stripe_width=160" --index 0 --mgsnode=miscmds7@tcp --fsname fs25 --verbose --reformat /dev/sda1
         Permanent disk data:
      Target:     fs25:OST0000
      Index:      0
      Lustre FS:  fs25
      Mount type: ldiskfs
      Flags:      0x62
                    (OST first_time update )
      Persistent mount opts: ,errors=remount-ro
      Parameters: mgsnode=141.34.22.70@tcp
      device size = 152591358MB
      formatting backing filesystem ldiskfs on /dev/sda1
          target name   fs25:OST0000
          kilobytes     156253550720
          options         -J size=1024 -I 512 -i 1048576 -O uninit_bg,extents,dir_nlink,quota,project,huge_file,64bit,^resize_inode,large_dir,^fast_commit,flex_bg -G 256 -E stride=16,stripe_width=160,lazy_journal_init,lazy_itable_init="0",packed_meta_blocks -F
      mkfs_cmd = mke2fs -j -b 4096 -L fs25:OST0000   -J size=1024 -I 512 -i 1048576 -O uninit_bg,extents,dir_nlink,quota,project,huge_file,64bit,^resize_inode,large_dir,^fast_commit,flex_bg -G 256 -E stride=16,stripe_width=160,lazy_journal_init,lazy_itable_init="0",packed_meta_blocks -F /dev/sda1 156253550720k
      cmd: mke2fs -j -b 4096 -L fs25:OST0000   -J size=1024 -I 512 -i 1048576 -O uninit_bg,extents,dir_nlink,quota,project,huge_file,64bit,^resize_inode,large_dir,^fast_commit,flex_bg -G 256 -E stride=16,stripe_width=160,lazy_journal_init,lazy_itable_init="0",packed_meta_blocks -F /dev/sda1 156253550720k
      mke2fs 1.47.1-wc2 (08-Nov-2024)
      detected raid stride 65536 too large, use optimum 512
      detected raid stripe width 655360 too large, use optimum 512
      /dev/sda1 contains a ext4 file system labelled 'fs25:OST0000'
          last mounted on /tmp/mntoarsDO on Thu Dec  5 15:05:12 2024
      Creating filesystem with 39063387680 4k blocks and 152591360 inodes
      Filesystem UUID: 87853c5b-29eb-41d8-bf23-f9860c025b28
      Superblock backups stored on blocks: 
          32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208, 4096000, 
          7962624, 11239424, 20480000, 23887872, 71663616, 78675968, 102400000, 
          214990848, 512000000, 550731776, 644972544, 1934917632, 2560000000, 
          3855122432, 5804752896, 12800000000, 17414258688, 26985857024
      Allocating group tables: done                            
      Writing inode tables: done                            
      Creating journal (262144 blocks): done
      Writing superblocks and filesystem accounting information: done           
      Writing CONFIGS/mountdata
      

      I am using stride=16,stripe_with_160 for a RAID6 on 12 drives with a strip size of 64k. So it appears the values are detected correctly by mke2fs, but, the recommended value of 512 is hard-coded.

      This was introduced in commit ad30685c9d1a23822f1072aa93d377a1688e6a21 to e2fsprogs for LU-12158 mke2fs: avoid too large stride and stripe_width

      Attachments

        Activity

          People

            wc-triage WC Triage
            waschk Götz Waschk
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

            Dates

              Created:
              Updated: