tried removing the daemon first, and that kinda blew up. ceph orch daemon rm --force mon.tempmon ceph orch host rm tempmon now there are two problems. 1. ceph is still looking for it, services: mon: 4 daemons, quorum ceph1,ceph2,ceph3 (age 3s), out of quorum: tempmon mgr: ceph1.oqptlg(active, since 9h), standbys: ceph2.hezrvv osd: 9 osds: 9 up (since 8h), 9 in (since 8h) 2. more worrying the cephadm module is failing somewhere INFO:cephadm:Inferring fsid 09e9711e-db88-11ea-b8c2-791b9888d2f2 INFO:cephadm:Using recent ceph image ceph/ceph:v15 cluster: id: 09e9711e-db88-11ea-b8c2-791b9888d2f2 health: HEALTH_ERR Module 'cephadm' has failed: must be str, not NoneType 1/4 mons down, quorum ceph1,ceph2,ceph3 root@ceph1:/home/vagrant# ceph health detail INFO:cephadm:Inferring fsid 09e9711e-db88-11ea-b8c2-791b9888d2f2 INFO:cephadm:Using recent ceph image ceph/ceph:v15 HEALTH_ERR Module 'cephadm' has failed: must be str, not NoneType; 1/4 mons down, quorum ceph1,ceph2,ceph3 [ERR] MGR_MODULE_ERROR: Module 'cephadm' has failed: must be str, not NoneType Module 'cephadm' has failed: must be str, not NoneType [WRN] MON_DOWN: 1/4 mons down, quorum ceph1,ceph2,ceph3 mon.tempmon (rank 3) addr [v2:10.16.16.10:3300/0,v1:10.16.16.10:6789/0] is down (out of quorum) is there another step that should be taken? Id expect "ceph orch host rm" to also take anything it managed out of the cluster. On Mon, Aug 10, 2020 at 12:42 PM pixel fairy <pixelfairy@gmail.com> wrote:
made a cluster of 2 osd hosts, and one temp monitor. then added another osd host and did a "ceph orch host rm tempmon". this all in vagrant (libvirt), with the generic/ubuntu2004 box.
INFO:cephadm:Inferring fsid 5426a59e-db33-11ea-8441-b913b695959d INFO:cephadm:Using recent ceph image ceph/ceph:v15 cluster: id: 5426a59e-db33-11ea-8441-b913b695959d health: HEALTH_WARN 2 stray daemons(s) not managed by cephadm 1 stray host(s) with 2 daemon(s) not managed by cephadm
added 2 more osd hosts, and ceph -s gave me this, services: mon: 6 daemons, quorum ceph5,ceph4,tempmon,ceph3,ceph2,ceph1 (age 33m) mgr: ceph5.erdofb(active, since 82m), standbys: tempmon.xkrlmm, ceph3.xjuecs osd: 15 osds: 15 up (since 33m), 15 in (since 33m)
my guess is cephadm wanted 5 managed mons, so did that, but still never removed the removed mon on the removed host. its still up. this is just a vagrant file. so i have two questions.
1. how do you remove that other host and its daemons from the cluster? 2. how would you recover from a host being destroyed?
p.s. tried google Your search - "ceph orch host rm" "stray daemons(s) not manage by cephadm" - did not match any documents.