Ask Your Question
1

trove guest instance starts normally, but trove-taskmanager still reports Error state

asked 2016-08-23 08:32:57 -0500

generalkalbasa gravatar image

Good day community! i have a multi node openstack mitaka deployment with secure management network and external network for instances as in openstack reference install guide.

controller has access to both networks. Trove-taskmanager uses rabbitMQ with rabbit_host set to CONTROLLER_MANAGEMENT_HOSTNAME, while guest agent uses rabbit_host set to CONTROLLER_EXTERNAL_IP.

I've have manualy built ubuntu 14.04 guest image for trove with mysql 5.6. my trove-guestagent.conf looks like this (have rabbit in both oslo and default just in case)

[DEFAULT]
verbose = True
debug = True
rabbit_host = CONTROLLER_EXTERNAL_IP
rabbit_userid = openstack
rabbit_password = PASS
nova_proxy_admin_user = admin
nova_proxy_admin_pass = ADMIN_PASS
nova_proxy_admin_tenant_name = service
trove_auth_url = http://CONTROLLER_MANAGEMENT_HOSTNAME:35357/v3
rpc_backend = rabbit
log_dir = /var/log/trove/
log_file = trove-guestagent.log
datastore_manager = mysql
datastore_registry_ext = mysql:trove.guestagent.datastore.mysql.manager.Manager

[oslo_messaging_rabbit]
rabbit_host = CONTROLLER_EXTERNAL_IP
rabbit_userid = openstack
rabbit_password = PASS

trove_auth_url is set to management host (though the instance has no access to that network), but accoroding to logs guest-agent doesn't even try to connect to that url so i don't think it is related to the issue.

When i start the instance the guest agent starts normaly, connects to rabbitmq and receives all the tasks it needs to do and finishes them without any error. in the end i have a fully functional Mysql database, accessible using the user/pass i entered during the creation. In the end, guestagent repeats the following message in the logs:

2016-08-23 12:53:44.685 1620 DEBUG oslo_service.periodic_task [-] Running periodic task Manager.update_status run_periodic_tasks /usr/local/lib/python2.7/dist-packages/oslo_service/periodic_task.py:215
2016-08-23 12:53:44.686 1620 DEBUG trove.guestagent.datastore.manager [-] Update status called. update_status /usr/lib/python2.7/dist-packages/trove/guestagent/datastore/manager.py:246
2016-08-23 12:53:44.687 1620 DEBUG trove.guestagent.datastore.service [-] Determining status of DB server. update /usr/lib/python2.7/dist-packages/trove/guestagent/datastore/service.py:178
2016-08-23 12:53:44.688 1620 DEBUG oslo_concurrency.processutils [-] Running cmd (subprocess): /usr/bin/mysqladmin ping execute /usr/local/lib/python2.7/dist-packages/oslo_concurrency/processutils.py:344
2016-08-23 12:53:44.737 1620 DEBUG oslo_concurrency.processutils [-] CMD "/usr/bin/mysqladmin ping" returned: 0 in 0.049s execute /usr/local/lib/python2.7/dist-packages/oslo_concurrency/processutils.py:374
2016-08-23 12:53:44.741 1620 INFO trove.guestagent.datastore.mysql_common.service [-] MySQL Service Status is RUNNING.
2016-08-23 12:53:44.742 1620 DEBUG trove.guestagent.datastore.service [-] Casting set_status message to conductor (status is 'running'). set_status /usr/lib/python2.7/dist-packages/trove/guestagent/datastore/service.py:162
2016-08-23 12:53:44.745 1620 DEBUG trove.conductor.api [-] Making async call to cast heartbeat for instance: 7a54fad1-b87f-484b-8ed0-a413f68cedad heartbeat /usr/lib/python2.7/dist-packages/trove/conductor/api.py:49
2016-08-23 12:53:44.747 1620 DEBUG oslo_messaging._drivers.amqpdriver [-] CAST unique_id: b496930502f5448f8b9864b892d3162c exchange 'openstack' topic 'trove-conductor' _send /usr/local/lib/python2.7/dist-packages/oslo_messaging/_drivers/amqpdriver.py:448
2016-08-23 12:53:44.752 1620 DEBUG trove.guestagent.datastore.service [-] Successfully cast set_status. set_status /usr/lib/python2.7/dist-packages/trove/guestagent/datastore/service.py:168

But on the controller ... (more)

edit retag flag offensive close merge delete

Comments

Could you please go through this thread Let us know the outcome.

Please also check

sunnyarora gravatar imagesunnyarora ( 2016-08-25 21:39:13 -0500 )edit

i have checked through those threads:1st one has another issue (stuck in build) 2nd one does have a kinda similar issue, but has no solution to it, unfortunately.

generalkalbasa gravatar imagegeneralkalbasa ( 2016-08-26 05:25:46 -0500 )edit

most trove issues are related to trove guest images not working properly, but mine is kinda working (according to logs and the ability to access mysql), but it somehow does not report the "success" to trove-taskmanager MQ

generalkalbasa gravatar imagegeneralkalbasa ( 2016-08-26 05:28:44 -0500 )edit

Look in your Trove Conductor log. Was the message from the guest successfully received and processed on the conductor? If you are running Mitaka, one issues around this time is caused by a mismatch between the versions of oslo.context on the guest and the controller.

amrith gravatar imageamrith ( 2016-08-26 07:20:12 -0500 )edit

trove conductor log from that day shows only regular INFO and DEBUG messages that it has started working at service restart long before the test instance launch. No messages with the timestamp close to the testing time

generalkalbasa gravatar imagegeneralkalbasa ( 2016-08-26 11:42:05 -0500 )edit

1 answer

Sort by ยป oldest newest most voted
0

answered 2017-02-04 13:54:52 -0500

AlexZ gravatar image

You need to add control_exchange = trove in your trove-guestagent.conf.

edit flag offensive delete link more

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower

Stats

Asked: 2016-08-23 08:32:57 -0500

Seen: 617 times

Last updated: Feb 04 '17