Hi, I'm building a lab with virtual machines. I build a set up with only 2 nodes, 2 osd per nodes and I have a host that use mount.cephfs Each 2 ceph nodes runs services mon + mgr + mds and has cephadm command. If I stop a node, all commands hang. Can't use dashboard, can't use ceph -s or any ceph command, and my cephfs on the third host stop to respond too (ex: with ls command) All come back, when I power on the stopped node Why is there no failover ?? Thanks you 'Jof
On 15.06.21 15:16, nORKy wrote:
Why is there no failover ??
Because only one MON out of two is not in the majority to build a quorum. Regards -- Robert Sander Heinlein Support GmbH Schwedter Str. 8/9b, 10119 Berlin https://www.heinlein-support.de Tel: 030 / 405051-43 Fax: 030 / 405051-19 Amtsgericht Berlin-Charlottenburg - HRB 93818 B Geschäftsführer: Peer Heinlein - Sitz: Berlin
Hi, That's right! We're currently evaluating a similar setup with two identical HW nodes (on two different sites), with OSD, MON and MDS each, and both nodes have CephFS mounted. The goal is to build a minimal self-contained shared filesystem that remains online during planned updates and can somehow survive should disaster strike at one of the two sites. We added a third node (a small VM) running only a monitor to avoid exactly the described problem. Best, Christoph On 15/06/2021 15.32, Robert Sander wrote:
On 15.06.21 15:16, nORKy wrote:
Why is there no failover ??
Because only one MON out of two is not in the majority to build a quorum.
Regards
_______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io
-- Dr. Christoph Brüning Universität Würzburg HPC & DataManagement @ ct.qmat & RZUW Am Hubland D-97074 Würzburg Tel.: +49 931 31-80499
Hi, Thank you guys. I deployed a third monitor and failover works. Thanks you Le mar. 15 juin 2021 à 16:15, Christoph Brüning < christoph.bruening@uni-wuerzburg.de> a écrit :
Hi,
That's right!
We're currently evaluating a similar setup with two identical HW nodes (on two different sites), with OSD, MON and MDS each, and both nodes have CephFS mounted.
The goal is to build a minimal self-contained shared filesystem that remains online during planned updates and can somehow survive should disaster strike at one of the two sites.
We added a third node (a small VM) running only a monitor to avoid exactly the described problem.
Best, Christoph
On 15/06/2021 15.32, Robert Sander wrote:
On 15.06.21 15:16, nORKy wrote:
Why is there no failover ??
Because only one MON out of two is not in the majority to build a quorum.
Regards
_______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io
-- Dr. Christoph Brüning Universität Würzburg HPC & DataManagement @ ct.qmat & RZUW Am Hubland D-97074 Würzburg Tel.: +49 931 31-80499 _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io
Hi, On 15.06.21 16:15, Christoph Brüning wrote:
Hi,
That's right!
We're currently evaluating a similar setup with two identical HW nodes (on two different sites), with OSD, MON and MDS each, and both nodes have CephFS mounted.
The goal is to build a minimal self-contained shared filesystem that remains online during planned updates and can somehow survive should disaster strike at one of the two sites.
This sounds like a use case for DRBD, maybe with OCFS2 on top as cluster(ed) filesystem. Ceph is overkill, and not really suited for two hosts setups. Regards, Burkhard
This also sounds like a possible GlusterFS use case. Regards, -Jamie On Tue, Jun 15, 2021 at 12:30 PM Burkhard Linke < Burkhard.Linke@computational.bio.uni-giessen.de> wrote:
Hi,
On 15.06.21 16:15, Christoph Brüning wrote:
Hi,
That's right!
We're currently evaluating a similar setup with two identical HW nodes (on two different sites), with OSD, MON and MDS each, and both nodes have CephFS mounted.
The goal is to build a minimal self-contained shared filesystem that remains online during planned updates and can somehow survive should disaster strike at one of the two sites.
This sounds like a use case for DRBD, maybe with OCFS2 on top as cluster(ed) filesystem. Ceph is overkill, and not really suited for two hosts setups.
Regards,
Burkhard
_______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io
-- Jamie Fargen Senior Consultant jfargen@redhat.com 813-817-4430
It's easy. The problem ise OSD's are still up because there is not enough down mon_osd_min_down_reporters and due to this problem MDS is stucking. The solution is "mon_osd_min_down_reporters = 1" Due to "two node" cluster and "replicated 2" with "chooseleaf host" the reporter count should be set to 1 but on a malfunction this could be a serious problem. In LAB environment you're fine. nORKy <joff.au@gmail.com>, 15 Haz 2021 Sal, 16:18 tarihinde şunu yazdı:
Hi,
I'm building a lab with virtual machines.
I build a set up with only 2 nodes, 2 osd per nodes and I have a host that use mount.cephfs Each 2 ceph nodes runs services mon + mgr + mds and has cephadm command.
If I stop a node, all commands hang. Can't use dashboard, can't use ceph -s or any ceph command, and my cephfs on the third host stop to respond too (ex: with ls command) All come back, when I power on the stopped node
Why is there no failover ??
Thanks you
'Jof _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io
participants (6)
-
Burkhard Linke
-
Christoph Brüning
-
Jamie Fargen
-
mhnx
-
nORKy
-
Robert Sander