Project

General

Profile

Bug #11811

erasure-code: lrc chunks are not where they are supposed to be

Added by Loïc Dachary almost 9 years ago. Updated almost 9 years ago.

Status:
Rejected
Priority:
Normal
Assignee:
Category:
-
Target version:
-
% Done:

0%

Source:
other
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

  • hammer
  • MON=3 OSD=10 ./vstart.sh -X -n -l mon osd
  • profile=ecprofile
  • ceph osd erasure-code-profile set $profile pluing=lrc k=4 m=2 l=3 ruleset-failure-domain=osd
  • poolname=cephlrc
  • ceph osd pool create $poolname 1 1 erasure $profile
  • edit hw.cc to modify the location of the ceph.conf file
  • compile with
    g++ -g -c attachment:hw.cc -o hw.o ; g++ -g hw.o -lrados -o hw or 

    if ceph is installed from packages or
    g++ -I include -L .libs hw.cc -l rados -o hw 
    

    if running from sources
  • ./hw
  • verify that all chunks have the same size with ls -l dev/osd?/current/1.0*/
    dev/osd0/current/1.0s1_head/:
    total 5008
    -rw-r--r-- 1 loic loic 5120000 mai   29 13:49 another2__head_02479CCC__1_ffffffffffffffff_1
    -rw-r--r-- 1 loic loic       0 mai   29 13:38 __head_00000000__1_ffffffffffffffff_1
    
    dev/osd2/current/1.0s2_head/:
    total 5008
    -rw-r--r-- 1 loic loic 5120000 mai   29 13:49 another2__head_02479CCC__1_ffffffffffffffff_2
    -rw-r--r-- 1 loic loic       0 mai   29 13:38 __head_00000000__1_ffffffffffffffff_2
    
    dev/osd3/current/1.0s4_head/:
    total 5008
    -rw-r--r-- 1 loic loic 5120000 mai   29 13:49 another2__head_02479CCC__1_ffffffffffffffff_4
    -rw-r--r-- 1 loic loic       0 mai   29 13:38 __head_00000000__1_ffffffffffffffff_4
    
    dev/osd7/current/1.0s5_head/:
    total 5008
    -rw-r--r-- 1 loic loic 5120000 mai   29 13:49 another2__head_02479CCC__1_ffffffffffffffff_5
    -rw-r--r-- 1 loic loic       0 mai   29 13:38 __head_00000000__1_ffffffffffffffff_5
    
    dev/osd8/current/1.0s0_head/:
    total 5008
    -rw-r--r-- 1 loic loic 5120000 mai   29 13:49 another2__head_02479CCC__1_ffffffffffffffff_0
    -rw-r--r-- 1 loic loic       0 mai   29 13:38 __head_00000000__1_ffffffffffffffff_0
    
    dev/osd9/current/1.0s3_head/:
    total 5008
    -rw-r--r-- 1 loic loic 5120000 mai   29 13:49 another2__head_02479CCC__1_ffffffffffffffff_3
    -rw-r--r-- 1 loic loic       0 mai   29 13:38 __head_00000000__1_ffffffffffffffff_3
    

hw.cc View - hw.cc (3.13 KB) Loïc Dachary, 05/29/2015 11:10 AM

History

#1 Updated by Loïc Dachary almost 9 years ago

#2 Updated by Loïc Dachary almost 9 years ago

  • Description updated (diff)
  • Status changed from New to In Progress

#3 Updated by Loïc Dachary almost 9 years ago

  • Description updated (diff)
  • Status changed from In Progress to Need More Info

Shylesh, do you observe the same results when you run your tests ?

#4 Updated by Loïc Dachary almost 9 years ago

  • Description updated (diff)

#5 Updated by Loïc Dachary almost 9 years ago

  • Description updated (diff)

#6 Updated by Loïc Dachary almost 9 years ago

  • Assignee changed from Loïc Dachary to Shylesh Kumar

#7 Updated by Loïc Dachary almost 9 years ago

  • Status changed from Need More Info to Rejected

The confusion came from the fact that the content of some shards is full of NULL, as expected.

Also available in: Atom PDF