Project

General

Profile

Bug #8184

krbd: make sure we have latest osdmap on 'rbd map'

Added by Ilya Dryomov almost 10 years ago. Updated almost 8 years ago.

Status:
Resolved
Priority:
Normal
Assignee:
Target version:
-
% Done:

0%

Source:
other
Tags:
Backport:
Regression:
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

$ ceph osd pool create pool1 8 8
$ rbd create --size 1000 pool1/img1
$ sudo rbd map pool1/img1
$ ceph osd pool create pool2 8 8
$ rbd create --size 1000 pool2/img2
$ sudo rbd map pool2/img2
rbd: sysfs write failed
rbd: map failed: (2) No such file or directory

syslog (148 KB) Luciano Andrade, 05/28/2016 02:54 PM

History

#1 Updated by Ilya Dryomov almost 10 years ago

  • Assignee set to Ilya Dryomov

An attempt is in wip-rbd-maposdmap, Sage suggested the mon_get_version approach.

#2 Updated by Ilya Dryomov almost 10 years ago

  • Status changed from In Progress to Fix Under Review

wip-rbd-maposdmap-v2;
"rbd: make sure we have latest osdmap on 'rbd map'" on ceph-devel.

#3 Updated by Ilya Dryomov almost 10 years ago

  • Status changed from Fix Under Review to In Progress

#4 Updated by Ilya Dryomov almost 10 years ago

  • Status changed from In Progress to Fix Under Review

#5 Updated by Ilya Dryomov almost 10 years ago

  • Status changed from Fix Under Review to Resolved

#6 Updated by Jeff Epstein about 9 years ago

I've been experiencing this problem recently.

root@coordinator:~# cat /proc/version
Linux version 3.18.4pl2 (root@lab) (gcc version 4.7.2 (Debian 4.7.2-5) ) #3 SMP Thu Jan 29 21:11:23 CET 2015
root@coordinator:~# ceph -v
ceph version 0.87 (c51c8f9d80fa4e0168aa52685b8de40e42758578)

And then:

root@host3:/# rbd -p ceph:xqxcvipfv73gw--git ls
ceph:xqxcvipfv73gw--git
root@host3:/# rbd -p ceph:xqxcvipfv73gw--git map ceph:xqxcvipfv73gw--git
rbd: sysfs write failed
rbd: map failed: (2) No such file or directory

Is this a possible regression?

#7 Updated by Ilya Dryomov about 9 years ago

No, I don't think so. Can you do

echo 9 >/proc/sysrq-trigger
echo 'module libceph +p' >/sys/kernel/debug/dynamic_debug/control
echo 'module rbd +p' >/sys/kernel/debug/dynamic_debug/control
rbd -p ceph:xqxcvipfv73gw--git map ceph:xqxcvipfv73gw--git
echo 'module libceph -p' >/sys/kernel/debug/dynamic_debug/control
echo 'module rbd -p' >/sys/kernel/debug/dynamic_debug/control

and attach the syslog?
What do you mean by recently? Did it coincide with ceph upgrade or kernel upgrade?

#8 Updated by Luciano Andrade almost 8 years ago

I run

echo 9 >/proc/sysrq-trigger
echo 'module libceph +p' >/sys/kernel/debug/dynamic_debug/control
echo 'module rbd +p' >/sys/kernel/debug/dynamic_debug/control
echo 'module libceph -p' >/sys/kernel/debug/dynamic_debug/control
echo 'module rbd -p' >/sys/kernel/debug/dynamic_debug/control

and then

sudo rbd map foo --pool rbd --name client.admin -m ceph-server-1

I think the relevant part of the error is

[18842.653363] SGI XFS with ACLs, security attributes, realtime, large block/inode numbers, no debug enabled
[18843.403297] xor: automatically using best checksumming function:
[18843.455759]    avx       : 24724.000 MB/sec
[18843.619217] raid6: sse2x1    7724 MB/s
[18843.688345] raid6: sse2x2   10194 MB/s
[18843.756344] raid6: sse2x4   11351 MB/s
[18843.756451] raid6: using algorithm sse2x4 (11351 MB/s)
[18843.756500] raid6: using ssse3x2 recovery algorithm
[18844.197126] bio: create slab <bio-1> at 1
[18844.200025] Btrfs loaded
[20597.729441] Key type ceph registered
[20597.729515] libceph: loaded (mon/osd proto 15/24)
[20597.749503] rbd: loaded rbd (rados block device)
[20597.752957] libceph: mon0 172.21.12.12:6789 feature set mismatch, my 4a042a42 < server's 2004a042a42, missing 20000000000
[20597.754058] libceph: mon0 172.21.12.12:6789 socket error on read
[20607.748535] libceph: mon0 172.21.12.12:6789 feature set mismatch, my 4a042a42 < server's 2004a042a42, missing 20000000000
[20607.749429] libceph: mon0 172.21.12.12:6789 socket error on read
[20617.754257] libceph: mon0 172.21.12.12:6789 feature set mismatch, my 4a042a42 < server's 2004a042a42, missing 20000000000
[20617.755252] libceph: mon0 172.21.12.12:6789 socket error on read
[20627.758599] libceph: mon0 172.21.12.12:6789 feature set mismatch, my 4a042a42 < server's 2004a042a42, missing 20000000000
[20627.759494] libceph: mon0 172.21.12.12:6789 socket error on read
[20637.763553] libceph: mon0 172.21.12.12:6789 feature set mismatch, my 4a042a42 < server's 2004a042a42, missing 20000000000
[20637.765272] libceph: mon0 172.21.12.12:6789 socket error on read
[20647.768654] libceph: mon0 172.21.12.12:6789 feature set mismatch, my 4a042a42 < server's 2004a042a42, missing 20000000000
[20647.769914] libceph: mon0 172.21.12.12:6789 socket error on read

The full syslog is attached.

#9 Updated by Ilya Dryomov almost 8 years ago

Your kernel is missing support for firefly tunables, see http://docs.ceph.com/docs/master/rados/operations/crush-map/#which-client-versions-support-crush-tunables3.
You can either upgrade your kernel or set your tunables to legacy values with "ceph osd crush tunables legacy".

Also available in: Atom PDF