Ceph config for radosgw doesnot get updated and persistent for admin user

Bug #1884811 reported by Yatindra Shashi on 2020-06-23
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
StarlingX
Undecided
Unassigned

Bug Description

Brief Description
-----------------
 Enabling Swift object storage in Openstack as per the mailing list email " http://lists.starlingx.io/pipermail/starlingx-discuss/2019-December/007136.html", but failed in duplex setup but succeded in simplex.

Steps to Reproduce
------------------
Repeat the steps suggested in "http://lists.starlingx.io/pipermail/starlingx-discuss/2019-December/007136.html"

Expected Behavior
------------------
Able to create swift container and upload the files..

Actual Behavior
----------------
- Fails multiple time with stx-openstack application apply but succeeded after deletign osh-openstack-ceph-rgw helm chart
- Authorization fails while trying to access Swift object storage->container from GUI

Reproducibility
---------------
Yes,

System Configuration
--------------------
Duplex, STX 3.0

Timestamp/Logs
--------------

---Sysinv.log>>
2020-06-22 16:46:49.536 2459 DEBUG armada.handlers.wait [-] [chart=openstack-ceph-rgw]: job ceph-ks-service is ready! handle_resource /usr/local/lib/python3.6/dist-packages/armada/handlers/wait.py:258^[[00m
2020-06-22 16:46:49.536 2459 DEBUG armada.handlers.wait [-] [chart=openstack-ceph-rgw]: job swift-ks-user not ready: Waiting for job swift-ks-user to be successfully completed... handle_resource /usr/local/lib/python3.6/dist-packages/armada/handlers/wait.py:260^[[00m
2020-06-22 16:47:13.332 2459 DEBUG armada.handlers.lock [-] Updating lock update_lock /usr/local/lib/python3.6/dist-packages/armada/handlers/lock.py:176^[[00m
2020-06-22 16:48:13.406 2459 DEBUG armada.handlers.lock [-] Updating lock update_lock /usr/local/lib/python3.6/dist-packages/armada/handlers/lock.py:176^[[00m
2020-06-22 16:49:13.505 2459 DEBUG armada.handlers.lock [-] Updating lock update_lock /usr/local/lib/python3.6/dist-packages/armada/handlers/lock.py:176^[[00m
2020-06-22 16:50:13.574 2459 DEBUG armada.handlers.lock [-] Updating lock update_lock /usr/local/lib/python3.6/dist-packages/armada/handlers/lock.py:176^[[00m
2020-06-22 16:51:13.639 2459 DEBUG armada.handlers.lock [-] Updating lock update_lock /usr/local/lib/python3.6/dist-packages/armada/handlers/lock.py:176^[[00m
2020-06-22 16:51:48.542 2459 ERROR armada.handlers.wait [-] [chart=openstack-ceph-rgw]: Timed out waiting for jobs (namespace=openstack, labels=(release_group=osh-openstack-ceph-rgw)). These jobs were not ready=['ceph-ks-endpoints', 'swift-ks-user']^[[00m

>>> Then removed osh-openstack-ceph-rgw helm chart and re-applied helm chart, stx-openstack was applied successfully but was below error.

controller-0:/home/sysadmin# tail -f /var/log/keystone/keystone-all.log
2020-06-23 09:14:59.770 1188411 WARNING keystone.auth.plugins.core [req-e352368e-7c76-494e-affd-9fc7182cf17c - - - - -] Could not find user: swift.: UserNotFound: Could not find user: swift.
2020-06-23 09:14:59.816 1188425 WARNING keystone.auth.plugins.core [req-22d63c61-29c2-4995-b5e0-fc323e42cdca - - - - -] Could not find user: swift.: UserNotFound: Could not find user: swift.
2020-06-23 09:14:59.824 1188425 WARNING keystone.server.flask.application [req-22d63c61-29c2-4995-b5e0-fc323e42cdca - - - - -] Authorization failed. The request you have made requires authentication. from 192.168.204.2: Unauthorized: The request you have made requires authentication.
2020-06-23 09:14:59.829 1188411 WARNING keystone.server.flask.application [req-e352368e-7c76-494e-affd-9fc7182cf17c - - - - -] Authorization failed. The request you have made requires authentication. from 192.168.204.2: Unauthorized: The request you have made requires authentication.

Workaround
----------
With the help of Austin, needed to change the /etc/ceph/ceph.conf file as:

[client.radosgw.gateway]
rgw_gc_obj_min_wait = 600
rgw_gc_processor_period = 300
rgw_keystone_api_version = 3
rgw_keystone_token_cache_size = 0
rgw_keystone_admin_domain = Default
rgw_keystone_url = http://keystone-api.openstack.svc.cluster.local:5000
rgw_s3_auth_use_keystone = true
user = root
rgw_max_put_size = 53687091200
rgw_keystone_admin_password = xxxxxxxx
rgw_gc_max_objs = 977
rgw_keystone_admin_user = admin
rgw_frontends = civetweb port=192.168.204.1:7480
log_file = /var/log/radosgw/radosgw.log
rgw_keystone_admin_project = admin
host = controller-0
rgw_keystone_accepted_roles = admin,_member_
rgw_gc_processor_max_time = 300
keyring = /etc/ceph/ceph.client.radosgw.gateway.keyring

Yatindra Shashi (yshashi) wrote :

- this configuration changed in the /etc/ceph/ceph.conf get to previous state after lock/unlock

Ghada Khalil (gkhalil) on 2020-06-25
tags: added: stx.storage
Ghada Khalil (gkhalil) wrote :

Adding the domain & release tag. This will require further investigation by the storage team.

tags: added: stx.3.0
To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers