Description of problem: The calamari api /api/v2/cluster/<fsid>/server returns OSD server names with crush bucket names appended. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. Create a cluster with say 1 mon and 1 osd node 2. Check the calamari api mentioned above (/api/v2/cluster/<fsid>/server) 3. Actual results: The output looks like below ----------------------------- [ { "fqdn": "dhcp46-150.lab.eng.blr.redhat.com-general", "hostname": "dhcp46-150.lab.eng.blr.redhat.com-general", "services": [ { "fsid": "f32fc340-779c-4233-bec3-d3a3ba354864", "type": "osd", "id": "0", "running": true } ], "frontend_addr": "10.70.46.150", "backend_addr": "10.70.46.150", "frontend_iface": null, "backend_iface": null, "managed": false, "last_contact": null, "boot_time": null, "ceph_version": null }, { "fqdn": "dhcp46-181.lab.eng.blr.redhat.com", "hostname": "dhcp46-181.lab.eng.blr.redhat.com", "services": [ { "fsid": "f32fc340-779c-4233-bec3-d3a3ba354864", "type": "mon", "id": "dhcp46-181", "running": true }, { Expected results: The hostname and fqdn for the OSD nodes should not have appended "-general" Additional info: The snippet from de-compiled crushmap is as below ------------------------------------------------- # buckets host dhcp46-150 { id -2 # do not change unnecessarily # weight 0.000 alg straw hash 0 # rjenkins1 } root default { id -1 # do not change unnecessarily # weight 0.000 alg straw hash 0 # rjenkins1 item dhcp46-150 weight 0.000 } host dhcp46-150.lab.eng.blr.redhat.com-general { id -3 # do not change unnecessarily # weight 0.000 alg straw hash 0 # rjenkins1 item osd.0 weight 0.000 } root general { id -4 # do not change unnecessarily # weight 0.000 alg straw hash 0 # rjenkins1 item dhcp46-150.lab.eng.blr.redhat.com-general weight 0.000 } ---------------------------
This causes import cluster flow broken in USM...
I can fix this. It's not going to be easy. I'll try to have it in by the 6.13 build
https://github.com/ceph/calamari/commit/d41bb9128b3548f94d2e1531f593109d061f317c
Hope this change returns proper fqdn. The same applicable for /api/v2/cluster/<fsid>/osd api as well
Gregory all the FQDN names should have domain names properly populated and not like dhcp46-239. Rather it should be shown as say "dhcp46-239.lab.eng.blt.redhat.com" ..
Yes this will provide a proper FQDN for /server and /osd
fqdn is displayed properly for the mon where calamari-lite is running. For rest of the nodes, it displays only host shortname. output of /api/v2/cluster/<fsid>/server will be attached. Tested on: calamari-server-1.4.1-1.el7cp.x86_64 ceph version 10.2.2-1.el7cp
Created attachment 1168309 [details] fqdn showing shortname
Harish the original issue in this BZ was fixed. We've got a new issue that captures what you observed: https://bugzilla.redhat.com/show_bug.cgi?id=1346207 Would you please file new BZs where there are issues that are identified around the initial issue?
Find this working with build calamari-server-1.4.2-1.el7cp.x86_64
I will be creating a separate defect for the issue found in comment 10 For now, based on comment 13, moving this defect to verified state.
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, 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://rhn.redhat.com/errata/RHBA-2016-1755.html