Bug #23286
closedmgr: ActivePyModules::list_servers_python() returns mgr with empty hostname
0%
Description
When running a vstart.sh cluster on master at cf52fc5a, list_servers_python()
returns:
[ { "services": [ { "type": "mgr", "id": "x" } ], "hostname": "", "ceph_version": "" }, { "services": [ { "type": "mds", "id": "a" }, { "type": "mds", "id": "b" }, { "type": "mds", "id": "c" }, { "type": "mon", "id": "a" }, { "type": "mon", "id": "b" }, { "type": "mon", "id": "c" }, { "type": "osd", "id": "0" }, { "type": "osd", "id": "1" }, { "type": "osd", "id": "2" }, { "type": "rgw", "id": "rgw" } ], "hostname": "ubuntu", "ceph_version": "ceph version 13.0.1-2832 .. mimic (dev)" } ]
This breaks the dashboard_v2 Host tab:
This is ceph mgr dump
:
$ ceph mgr dump | jq . 2018-03-09 11:48:27.125 7fe28a849700 -1 WARNING: all dangerous and experimental features are enabled. 2018-03-09 11:48:27.177 7fe28a849700 -1 WARNING: all dangerous and experimental features are enabled. { "epoch": 8, "active_gid": 4137, "active_name": "x", "active_addr": "192.168.178.29:6800/9933", "available": true, "standbys": [], "modules": [ "balancer", "dashboard", "dashboard_v2", "restful", "status" ], "available_modules": [ { "name": "balancer", "can_run": true, "error_string": "" }, { "name": "dashboard", "can_run": true, "error_string": "" }, { "name": "dashboard_v2", "can_run": true, "error_string": "" }, { "name": "influx", "can_run": false, "error_string": "influxdb python module not found" }, { "name": "localpool", "can_run": true, "error_string": "" }, { "name": "prometheus", "can_run": true, "error_string": "" }, { "name": "restful", "can_run": true, "error_string": "" }, { "name": "selftest", "can_run": true, "error_string": "" }, { "name": "smart", "can_run": true, "error_string": "" }, { "name": "status", "can_run": true, "error_string": "" }, { "name": "zabbix", "can_run": true, "error_string": "" } ], "services": { "dashboard": "http://ubuntu:41116/", "dashboard_v2": "http://ubuntu:43116/", "restful": "https://ubuntu:42116/" } }
Updated by Chang Liu about 6 years ago
- Assignee set to Chang Liu
- Hello, what's your ceph version? -
Oh, sorry. it's "ceph version 13.0.1-2832 .. mimic (dev)".
Updated by John Spray about 6 years ago
- Status changed from New to 12
This bug is a result of how we populate the mgrs into DaemonState from DaemonServer::got_mgr_map without ever reading the mgr daemon metadata from the monitors. We need to do something similar to what is done for mon/osd/mds daemons from the Mgr class (MetadataUpdate etc).
Updated by Jan Fajerski about 6 years ago
I have a potential fix (updating the metadata in DaemonServer::got_mgr_map()) but encountered http://tracker.ceph.com/issues/23330.
Updated by John Spray about 6 years ago
- Related to Bug #23330: mon command "mgr metadata $name" has inconsistent argument naming added
Updated by Nathan Cutler about 6 years ago
Updated by Jan Fajerski over 5 years ago
Was fixed here https://github.com/ceph/ceph/pull/20875
Updated by Nathan Cutler about 5 years ago
- Status changed from Closed to Pending Backport
- Pull request ID set to 20875
Updated by Nathan Cutler about 5 years ago
- Copied to Backport #39234: luminous: mgr: ActivePyModules::list_servers_python() returns mgr with empty hostname added
Updated by Sebastian Wagner over 4 years ago
- Related to Bug #41549: mgr: ActivePyModules::list_servers_python() returns mds with empty hostname added
Updated by Nathan Cutler about 3 years ago
- Status changed from Pending Backport to Resolved
While running with --resolve-parent, the script "backport-create-issue" noticed that all backports of this issue are in status "Resolved" or "Rejected".