Bug 1219344
| Summary: | 1.3.0: "ceph-deploy calamari connect <node>" fails | ||||||
|---|---|---|---|---|---|---|---|
| Product: | [Red Hat Storage] Red Hat Ceph Storage | Reporter: | Harish NV Rao <hnallurv> | ||||
| Component: | Build | Assignee: | Alfredo Deza <adeza> | ||||
| Status: | CLOSED ERRATA | QA Contact: | ceph-qe-bugs <ceph-qe-bugs> | ||||
| Severity: | urgent | Docs Contact: | |||||
| Priority: | unspecified | ||||||
| Version: | 1.3.0 | CC: | adeza, flucifre, gmeno, hyelloji, kdreyer, trhoden | ||||
| Target Milestone: | rc | ||||||
| Target Release: | 1.3.0 | ||||||
| Hardware: | x86_64 | ||||||
| OS: | Linux | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | ceph-deploy-1.5.24-1.el7cp | Doc Type: | Bug Fix | ||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2015-06-24 15:52:52 UTC | Type: | Bug | ||||
| Regression: | --- | Mount Type: | --- | ||||
| Documentation: | --- | CRM: | |||||
| Verified Versions: | Category: | --- | |||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||
| Embargoed: | |||||||
| Bug Depends On: | 1210038 | ||||||
| Bug Blocks: | |||||||
| Attachments: |
|
||||||
|
Description
Harish NV Rao
2015-05-07 06:40:52 UTC
The way this is supposed to work is that the cephdeploy.conf file should have a "[calamari-minion]" repository section. This is automatically created for the user when installing via the "ice_setup.py" script because that script takes care of creating paths and moving packages and dealing with repos. In that process it creates the section for the user (that you seem to be missing). In short: if the user is not going via the ice_setup.py, this is not supposed to work *unless* a "[calamari-minion]" repo section with all details needed exists in the cephdeploy.conf file. Alfredo, Your description in Comment #2 was true before, but it is no longer for Stockwell. As I mentioned in the email thread, there is no longer a "calamari-minions" repo needed at all. All the dependencies for a calamari minion are present in the ceph-mon *and* the ceph-osd channels. Therefore, ceph-deploy does not need to do any installation of repo files for calamari-minions, and ice-setup should create/host such a repo. ceph-deploy just needs to do the install as if the repo is already there. In that sense, this install *should* work without ice_setup being run. So ceph-deploy does need to modified such that it does not do any of those steps when using RHEL + RCHS. Basically, these lines need to be disabled for RHCS: https://github.com/ceph/ceph-deploy/blob/master/ceph_deploy/calamari.py#L23-L80 (In reply to Travis Rhoden from comment #3) > Therefore, ceph-deploy does not need to do any installation of repo files > for calamari-minions, and ice-setup should create/host such a repo. > ceph-deploy just needs to do the install as if the repo is already there. > Just to clarify, I meant "ice-setup should **NOT** create/host such a repo..." changing the severity as this is blocking calamari UI testing. Thanks Travis! I will try the work around. I am also facing the same problem as mentioned by Hemanth after applying the workaround. I am getting following message in GUI: "5 Ceph servers are connected to Calamari, but no Ceph cluster has been created yet. Please use ceph-deploy to create a cluster; please see the Inktank Ceph Enterprise documentation for more details." Calamari and Admin node IP: http://10.8.128.6 user = TEST12, password = ghj checking the calamari host I see that the ceph module has not been sync'd to the minions via salt.
I'll continue to investigate why.
[cephuser@Admin ~]$ sudo salt-key -L
Accepted Keys:
Admin
magna029
magna033
magna040
magna086
Unaccepted Keys:
Rejected Keys:
[cephuser@Admin ~]$ sudo salt '*' ceph.get_heartbeats
magna086:
'ceph.get_heartbeats' is not available.
magna029:
'ceph.get_heartbeats' is not available.
magna033:
'ceph.get_heartbeats' is not available.
magna040:
'ceph.get_heartbeats' is not available.
Admin:
----------
- boot_time:
1430872612
- ceph_version:
0.94.1-8.el7cp
- services:
----------
----------
The issue here is that the cluster is running a different version of salt from EPEL. This is not a supported setup. I recommend downgrading salt-minion and re-trying. gmeno$ ssh shadow_man.redhat.com Last login: Wed May 6 14:18:15 2015 from 10.8.128.2 [shadow_man@osd1 ~]$ salt-minion --version salt-minion 2014.7.5 (Helium) [shadow_man@osd1 ~]$ [shadow_man@osd1 ~]$ yum repolist Loaded plugins: priorities, product-id, subscription-manager Repo rhel-7-server-extras-rpms forced skip_if_unavailable=True due to: /etc/pki/entitlement/5269745257163545448-key.pem Repo rhel-7-server-rpms forced skip_if_unavailable=True due to: /etc/pki/entitlement/5269745257163545448-key.pem Repo rhel-7-server-optional-rpms forced skip_if_unavailable=True due to: /etc/pki/entitlement/5269745257163545448-key.pem lab-extras 10/10 rhel-7-fcgi-ceph 3/3 rhel-7-server-ceph-1.3-calamari-test 15/15 rhel-7-server-ceph-1.3-installer-test 1/1 rhel-7-server-ceph-1.3-mon-test 26/26 rhel-7-server-ceph-1.3-osd-test 23/23 repo id repo name status!epel/x86_64 Extra Packages for Enterprise Linux 7,819 !lab-extras lab-extras 10 !rhel-7-fcgi-ceph RHEL 7 Local fastcgi Repo 2+1 rhel-7-server-ceph-1.3-calamari-test Server-RH7-CEPH-CALAMARI-1.3 15 rhel-7-server-ceph-1.3-installer-test Server-RH7-CEPH-INSTALLER-1.3 1 rhel-7-server-ceph-1.3-mon-test Server-RH7-CEPH-MON-1.3 26 rhel-7-server-ceph-1.3-osd-test Server-RH7-CEPH-OSD-1.3 23 rhel-7-server-extras-rpms/x86_64 Red Hat Enterprise Linux 7 Server - Extras (RPMs) 52 rhel-7-server-optional-rpms/7Server/x86_64 Red Hat Enterprise Linux 7 Server - Optional (RPMs) 5,804 rhel-7-server-rpms/7Server/x86_64 Red Hat Enterprise Linux 7 Server (RPMs) 6,990 repolist: 20,742 [shadow_man@osd1 ~]$ yum info salt-minion Loaded plugins: priorities, product-id, subscription-manager Repo rhel-7-server-extras-rpms forced skip_if_unavailable=True due to: /etc/pki/entitlement/5269745257163545448-key.pem Repo rhel-7-server-rpms forced skip_if_unavailable=True due to: /etc/pki/entitlement/5269745257163545448-key.pem Repo rhel-7-server-optional-rpms forced skip_if_unavailable=True due to: /etc/pki/entitlement/5269745257163545448-key.pem Installed Packages Name : salt-minion Arch : noarch Version : 2014.7.5 Release : 1.el7 Size : 28 k Repo : installed From repo : epel Summary : Client component for Salt, a parallel remote execution system URL : http://saltstack.org/ License : ASL 2.0 Description : The Salt minion is the agent component of Salt. It listens for instructions : from the master, runs jobs, and returns results back to the master. [shadow_man@osd1 ~]$ I did following on all the nodes that had different salt-minion version:
1. Disabled the EPEL:
a) sudo vi /etc/yum.repos.d/epel.repo
b) change enabled=1 to enabled=0
2. installed correct version of salt-minion:
a) sudo yum distro-sync salt\*
3. sudo service salt-minion restart (without this GUI did not show dashboard)
Then opened calamari UI http://10.8.128.6.
It showed the Dashboard!!! Able to see the details of Ceph cluster!!!
Thanks a lot for the quick response and help.
Regards,
Harish
ceph-deploy calamari connect <node> executed successfully with the new ISO puddle generated on 26th May.. Moving to verified state. --------------------------------------------------------- Output of ceph-deploy calamari connect <MON_Node> http://pastebin.test.redhat.com/286920 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://access.redhat.com/errata/RHBA-2015:1183 |