daemon fails because cluster has different name than 'ceph'

[mkudlej]

It seems that ceph integration daemon work only if there is cluster with name ceph. It should work also if node is part of cluster with different name than ceph.
Example of call which breaks daemon:

$ python /usr/bin/ceph version -f json
Error initializing cluster client: Error('error calling conf_read_file: error code 22',)

which is part of start of daemon

systemctl start tendrl-cephd.service

[mkudlej]

This problem is worst than I expected. Command:

python /usr/bin/ceph version -f json

cannot be run on OSD nodes because they are not Ceph clients by default. So I think ceph command should be substitute by something else.
--> moving this issue to blockers

[shtripat]

Sent #118 to fix this.
We rather should be using ceph --version command which works on both mon and ceph nodes properly.

[sankarshanmukhopadhyay]

@shtripat is this fix planned to be available in the latest build?

[nthomas-redhat]

@sankarshanmukhopadhyay This was merged yesterday, so should be available in the last nightly build(1.2-02_21_2017_06_35_02 )

[sankarshanmukhopadhyay]

ACK. Thank you @nthomas-redhat Subsequent to testing, if found fixed, I'd request the reporter to mark this as closed. And similarly for other issues which have been resolved in the latest builds.

[sankarshanmukhopadhyay]

@mkudlej - once you commence testing, please verify if the fix works and we can resolve this issue.

[mkudlej]

With today packages I see properly stored ceph cluster in Etcd. Once I can confirm this by importing cluster with different name than ceph, I'll close this issue. This is blocked by Tendrl/api#82

[sankarshanmukhopadhyay]

Since Tendrl/api#82 seems to have been addressed, can we attempt to confirm the fix this week?

[mkudlej]

There is new blocker https://github.com/Tendrl/performance-monitoring/issues/58 which blocks also this.

[sankarshanmukhopadhyay]

The fix for https://github.com/Tendrl/performance-monitoring/issues/58 has been merged. Please check the latest builds when available and update this issue.

[mkudlej]

I don't see this issue with:

• api server:

tendrl-api-1.2.1-03_01_2017_01_51_02.noarch
tendrl-api-doc-1.2.1-03_01_2017_01_51_02.noarch
tendrl-api-httpd-1.2.1-03_01_2017_01_51_02.noarch
tendrl-dashboard-1.2.1-03_01_2017_13_39_11.noarch

• ceph node:

centos-release-ceph-jewel-1.0-1.el7.centos.noarch
ceph-base-10.2.3-0.el7.x86_64
ceph-common-10.2.3-0.el7.x86_64
ceph-mon-10.2.3-0.el7.x86_64
ceph-selinux-10.2.3-0.el7.x86_64
libcephfs1-10.2.3-0.el7.x86_64
python-cephfs-10.2.3-0.el7.x86_64
tendrl-ceph-integration-1.2.1-03_01_2017_06_35_02.noarch
tendrl-commons-1.2.1-2.el7.centos.noarch
tendrl-node-agent-1.2.1-03_01_2017_00_47_35.noarch
tendrl-performance-monitoring-1.2.1-03_01_2017_03_02_02.noarch

--> CLOSE