Bug 1331881 - ceph-ansible does not use systemd on Xenial
Summary: ceph-ansible does not use systemd on Xenial
Alias: None
Product: Red Hat Storage Console
Classification: Red Hat
Component: ceph-installer
Version: 2
Hardware: Unspecified
OS: Unspecified
Target Milestone: ---
: 2
Assignee: Andrew Schoen
QA Contact: sds-qe-bugs
: 1332920 1333339 1333442 (view as bug list)
Depends On:
TreeView+ depends on / blocked
Reported: 2016-04-29 22:20 UTC by Ken Dreyer (Red Hat)
Modified: 2016-08-23 19:49 UTC (History)
7 users (show)

Fixed In Version: ceph-ansible-1.0.5-8.el7scon
Doc Type: Bug Fix
Doc Text:
Clone Of:
Last Closed: 2016-08-23 19:49:39 UTC
Target Upstream Version:

Attachments (Terms of Use)

System ID Priority Status Summary Last Updated
Red Hat Bugzilla 1333339 None None None Never
Red Hat Bugzilla 1333442 None None None Never
Red Hat Product Errata RHEA-2016:1754 normal SHIPPED_LIVE New packages: Red Hat Storage Console 2.0 2017-04-18 19:09:06 UTC

Internal Links: 1333339 1333442

Description Ken Dreyer (Red Hat) 2016-04-29 22:20:43 UTC
Description of problem:
When deploying Ceph on Xenial (which uses systemd), the /api/mon/configure endpoint hangs forever. The reason ceph-ansible hangs forever is that it is waiting on the "ceph-mon | collect admin and bootstrap keys" task.

The problem is that ceph-create-keys waits for keys to appear, and the keys do not appear because Ansible has not properly started the ceph-mon service with systemd.

Running `systemctl start ceph-mon@<hostname>` creates the keys and allows Ansible to proceed with the rest of the playbook to configure the monitor.

Version-Release number of selected component (if applicable):

How reproducible:

Steps to Reproduce:
1. Prepare Xenial machines to host Ceph
2. POST to the installer /api/mon/install , /api/osd/install endpoints
3. Verify the package installations succeeded on the cluster nodes.
4. POST to the /api/mon/configure endpoint
5. Check /api/tasks

Actual results:
The /api/mon/configure task hangs. (Running `systemctl start ceph-mon@<hostname>` on the monitor node(s) allows it to continue.)

Expected results:
The /api/mon/configure task succeeds without intervention.

Additional info:
I found this problem in the monitor config step, and we need ceph-ansible to use systemd on Xenial for all services: ceph-mon and ceph-osd.

Comment 2 Ken Dreyer (Red Hat) 2016-05-04 20:57:00 UTC
As a reminder, until this bug is fixed, if you see "/api/mon/configure" taking longer than 50 seconds, log into the monitor node and run

sudo systemctl start ceph-mon@$(hostname)

Comment 3 Andrew Schoen 2016-05-04 21:22:29 UTC
I've made a PR that fixes this issue here: https://github.com/ceph/ceph-ansible/pull/758

Comment 4 Ken Dreyer (Red Hat) 2016-05-05 15:31:56 UTC
*** Bug 1332920 has been marked as a duplicate of this bug. ***

Comment 5 Rachana Patel 2016-05-05 21:43:31 UTC
Adding mon to existing ceph cluster also fails and looks like due to this bug.
Once this bug is fixed we might re run that to confirm.

Comment 6 Alfredo Deza 2016-05-09 14:11:44 UTC
*** Bug 1332920 has been marked as a duplicate of this bug. ***

Comment 7 Andrew Schoen 2016-05-09 16:00:15 UTC
The PR has been merged upstream: https://github.com/ceph/ceph-ansible/pull/758

Comment 9 Andrew Schoen 2016-05-09 19:57:25 UTC
*** Bug 1333442 has been marked as a duplicate of this bug. ***

Comment 10 Andrew Schoen 2016-05-09 19:58:32 UTC
*** Bug 1333339 has been marked as a duplicate of this bug. ***

Comment 16 errata-xmlrpc 2016-08-23 19:49:39 UTC
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.


Note You need to log in before you can comment on or make changes to this bug.