[lustre-discuss] lfs getstripe incomplete output

John Bauer bauerj at iodoctors.com
Wed Apr 1 06:18:00 PDT 2026


All,

I am applying a 2 component striping, including specifying the osts, to 
a file with an lfs setstripe command.  When I issue the lfs getstripe 
command on the resulting file, the osts for the 2nd component are not 
displayed even though they were explicitly set. I  have written my own 
program to get the striping information using the llapi functions and it 
reports the explicitly set osts are in the 2nd component.  It would 
appear that lfs getstripe is not displaying osts in components that 
start beyond the file size, assuming those components have not been 
instantiated.  Is this the desired behavior.

John


#!/bin/bash -v

file=/hpelustre/e63414/test.dat

rm -f ${file}

lfs setstripe \
         -E 10M \
         --stripe-count=2 \
         --stripe-size=1M \
         --ost=3,4 \
         -E EOF \
         --stripe-count=2 \
         --stripe-size=4M \
         --ost=2,1 \
         ${file}

lfs getstripe ${file}

Resulting output:


lfs getstripe ${file}
/hpelustre/e63414/test.dat
   lcm_layout_gen:    2
   lcm_mirror_count:  1
   lcm_entry_count:   2
     lcme_id:             1
     lcme_mirror_id:      0
     lcme_flags:          init
     lcme_extent.e_start: 0
     lcme_extent.e_end:   10485760
       lmm_stripe_count:  2
       lmm_stripe_size:   1048576
       lmm_pattern:       raid0
       lmm_layout_gen:    0
       lmm_stripe_offset: 3
       lmm_objects:
       - 0: { l_ost_idx: 3, l_fid: [0x100030000:0x971473:0x0] }
       - 1: { l_ost_idx: 4, l_fid: [0x100040000:0x9711c9:0x0] }

     lcme_id:             2
     lcme_mirror_id:      0
     lcme_flags:          0
     lcme_extent.e_start: 10485760
     lcme_extent.e_end:   EOF
       lmm_stripe_count:  2
       lmm_stripe_size:   4194304
       lmm_pattern:       raid0
       lmm_layout_gen:    0
       lmm_stripe_offset: 2
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lustre.org/pipermail/lustre-discuss-lustre.org/attachments/20260401/24996767/attachment.htm>


More information about the lustre-discuss mailing list