Bug 2227600 - [EC 8+6 profile on 4 nodes]: crush: choose leaf with type = 0 may incorrectly map out osds
Summary: [EC 8+6 profile on 4 nodes]: crush: choose leaf with type = 0 may incorrectly...
Keywords:
Status: CLOSED ERRATA
Alias: None
Product: Red Hat Ceph Storage
Classification: Red Hat Storage
Component: RADOS
Version: 6.1
Hardware: Unspecified
OS: Unspecified
unspecified
medium
Target Milestone: ---
: 8.0
Assignee: Samuel Just
QA Contact: Pawan
URL:
Whiteboard:
Depends On:
Blocks: 2317218
TreeView+ depends on / blocked
 
Reported: 2023-07-31 04:50 UTC by Vidushi Mishra
Modified: 2024-11-25 08:59 UTC (History)
11 users (show)

Fixed In Version: ceph-19.1.1-8.el9cp
Doc Type: Enhancement
Doc Text:
.New MSR CRUSH rules for erasure encoded pools Multi-step-retry (MSR) is a new type of CRUSH rule for placing multiple OSDs on each host. With this enhancement, `crush-osds-per-failure-domain` and `crush-num-failure-domains` can now be specified for erasure-coded (EC) pools during their creation. These pools use newly introduced MSR CRUSH rules to place multiple OSDs within each failure domain. For example, 14 OSDs split across 4 hosts. For more information, see /architecture/con_core_ceph-erasure-coding
Clone Of:
Environment:
Last Closed: 2024-11-25 08:59:19 UTC
Embargoed:


Attachments (Terms of Use)


Links
System ID Private Priority Status Summary Last Updated
Ceph Project Bug Tracker 62214 0 None None None 2023-08-09 21:45:56 UTC
Github ceph ceph pull 55332 0 None Merged crush: add multistep retry rules 2024-05-03 17:09:42 UTC
Red Hat Issue Tracker RHCEPH-7106 0 None None None 2023-07-31 04:51:40 UTC
Red Hat Product Errata RHBA-2024:10216 0 None None None 2024-11-25 08:59:24 UTC

Description Vidushi Mishra 2023-07-31 04:50:49 UTC
Description of problem:


If all of the OSDs in a host are marked out, crush_choose_indep on a leaf bucket with recurse_to_leaf=1 (chooseleaf rather than choose) will populate out2 before the is_out check:


The motivating example was:

rule ecpool-86 {
id 86
type erasure
step set_chooseleaf_tries 5
step set_choose_tries 100
step take default class hdd
step choose indep 4 type host
step chooseleaf indep 4 type osd
step emit
}

If all of the OSDs in a host are marked out, crush_choose_indep on a leaf bucket with recurse_to_leaf=1 (chooseleaf rather than choose) will populate out2 before the is_out check:


https://tracker.ceph.com/issues/62213


Version-Release number of selected component (if applicable):

ceph version 17.2.6-70.el9cp 

How reproducible:


Steps to Reproduce:
1.
2.
3.

Actual results:


Expected results:


Additional info:

Comment 14 errata-xmlrpc 2024-11-25 08:59:19 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (Red Hat Ceph Storage 8.0 security, bug fix, and enhancement updates), and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2024:10216


Note You need to log in before you can comment on or make changes to this bug.