Ask Your Question

JackLin's profile - activity

2019-03-23 10:15:20 -0500 received badge  Famous Question (source)
2017-10-06 01:27:50 -0500 received badge  Famous Question (source)
2017-10-06 01:27:50 -0500 received badge  Notable Question (source)
2017-04-26 12:05:42 -0500 received badge  Teacher (source)
2017-02-22 13:59:26 -0500 received badge  Taxonomist
2017-02-19 07:39:27 -0500 asked a question Sahara/swiftclient/keystoneclient ClientException: Unauthorized. Check username, password and tenant name/id

Hi! I am now using openstack mitaka with sahara-api on my controller and sahara-engine on my network node.

I successed to launch a hadoop 2.7.0 cluster.
But when I wanted to execute a job, I encountered this kind of error.

It seems that swiftclient's authentication faiied.
But I can successfully use "swift stat" with admin username and password.
Is there any one encountered the same error as me or knows how to fix it. Thanks a lot!

2017-02-14 14:16:58 -0500 commented answer Heat stevedore.extension Could not load 'glance.image': No module named openstack.common.apiclient

But I encountered another problem. :-C

2017-02-14 14:15:51 -0500 commented answer Heat stevedore.extension Could not load 'glance.image': No module named openstack.common.apiclient

Thanks very much!

2017-02-14 14:15:01 -0500 asked a question Heat Scaling_policy alarm_url is null

I am now using mitaka and installed manually.
After installed Heat, I wanted to try auto-scaling function but faced some problems.
Here is my yaml file.

And when I create the stack, it failed right after scaling_policy creating complete and OS::Ceilometer::alarm initiating

After researching for a while, I found that it might caused by the scaling_policy didn't create the alarm_url,
and the scaling_policy resource's alarm_url is null
(or what I guessed is wrong)

My openstack-heat-api-cfn.service works well, and I have create the endpoint of it.

Is there anyone know how can I fixed it or why it happened?
Thanks a lot!

2017-02-14 10:18:46 -0500 received badge  Notable Question (source)
2017-02-14 09:50:24 -0500 received badge  Popular Question (source)
2017-02-13 07:40:49 -0500 asked a question Heat stevedore.extension Could not load 'glance.image': No module named openstack.common.apiclient

I have encountered a problem after installing Heat.
The log said:
" stevedore.extension [-] Could not load 'glance.image': No module named openstack.common.apiclient "

Thus, when i use heat to launch an instance with image, the error will occur like below.

But actually, I have installed glanceclient and openstackclient and can use "glance image-list" to list all my image.
I have no idea how can I fix it. Is there any one encountered the same problem as me or know how to fix it? Thanks!

2017-01-23 10:29:06 -0500 answered a question Heat Autoscaling Where is OS::Ceilometer/Aodh::Alarm

Hi! Thanks for replying me.
I have found the answer of my question after tracing code.
Here is my heat source code and directory.
You can see that it use ceilometer not aodh as resources in my version which is Mitaka on CentOS.

But, In my /etc/heat/environment.d/default.yaml file

I don't know why it changed Ceilometer to Aodh in resource_registry.
After commenting out those line, OS::Ceilometer::Alarm appeared successfully.

2017-01-22 00:38:55 -0500 received badge  Famous Question (source)
2016-11-18 03:34:19 -0500 received badge  Notable Question (source)
2016-11-16 17:07:39 -0500 received badge  Famous Question (source)
2016-11-06 10:53:50 -0500 received badge  Popular Question (source)
2016-11-04 15:37:54 -0500 received badge  Popular Question (source)
2016-11-03 04:03:37 -0500 asked a question magnum integration with ceilometer


I have some questions of Ceilometer and Magnum.
I am using mitaka now with ceilometer and magnum installed.

According to these blueprints

Magnum emits notifications to Notification Bus and Ceilometer catches the notification.
But when I command "ceilometer meter-list", there's no any meter about magnum's bay info.
I want to ask that do I need to configure some other things in order to see the meters.

Thanks a lot!

2016-10-26 12:38:27 -0500 received badge  Student (source)
2016-10-25 08:16:35 -0500 asked a question Heat Autoscaling Where is OS::Ceilometer/Aodh::Alarm

Hi! I am using Mitaka now, and want to try the autoscaling.
I followed this tutorial
but found that in my version(heat:1.0.0) there is no resource type naming "OS::Ceilometer::Alarm" or "OS::Aodh::Alarm"
So where can I found the resource type or how can I create the resource type.

Thanks a lot!

2016-10-25 04:55:23 -0500 asked a question [Heat]keystoneauth1.exceptions.connection.ConnectFailure: Unable to establish connection to http://controller:8004/v1/...

I am using mitaka now, and have installed heat.
But I suffer from the error when "heat stack-list" command

here is the debug information

My endpoint is all right and heat service works fine like below

I can curl nova's endpoint , but it showed "connection refused" when curling the endpoint of heat.

Does anyone encounter the same problem or know what mistake I made.
Thanks a lot.

2016-10-12 07:12:10 -0500 asked a question magnum integrates with senlin and ceilometer


I have found an impressive demo on 2015 Tokyo Summit about Magnum integrating with Senlin and Ceilometer to do the autoscaling on both pod and VM level.

Here is the video (

I have install openstack mitaka with magnum, senlin and ceilometer.
I want to reproduce the demo but have some questions.
First, the demo said that magnum will use senlin to create a kubernetes bay, but actually magnum used Heat template to create kubernetes bay, so how can I make magnum use senlin to create the cluster?
Second, it seems that ceilometer doesn't integrate with magnum or kubernetes cAdvisor/Heapster, so how can ceilometer retrieve the resource usage information from the kubernetes bay?

Thanks very much!
Make openstack have the ability to automatically manage the container cluster on both VM and container level is so fatastic.

2016-09-29 00:32:50 -0500 commented question Magnum vm instance can ping controller ip but can't resolve it by hostname to notify heat of status

I got the same error as you. Have you solved the problem? Please let me know. Thanks!

2016-09-28 09:03:10 -0500 received badge  Editor (source)
2016-09-28 04:32:24 -0500 asked a question Magnum and Heat create bay failed key already exists, we are the master

Hi! I am now using openstack mitaka. I have installed magnum and heat, but create bay failed. Heat resource showed that kube_maaster CREATE_FAILED and the reason is CREATE aborted.

Here is the resource-show

The troubleshooting doc suggest me to log in to the VM to see the error

and here is the information form journalctl

Can anyone help me? Thanks a lot!


After installing the openstack-heat-api-cloudwatch and recreate the bay, I got the following error

and this is the information in the vm journalctl.


I found that I forgot to add docker-volumn-size After add this to baymodel and recreate the bay, It hung in here

2016-09-26 04:01:44 -0500 answered a question Magnum create bay failed

Hi! I have solved the problem.

The error was not in the magnum or heat, but in the mariaDB which openstack uses. The default maximum connection is too low that when too many services try to connect the database, it will raise the "too many connection" error and make api of nova, neutron...etc failed. This made my openstack so unstable and errors like HTTPInternalServerError appeared.

What I did was just changed the mariadb maximum connection in the configuration file and restarted it and all works fine now.

Below is the reference I found on google about how to solve the problem (

2016-09-23 09:07:45 -0500 received badge  Notable Question (source)
2016-09-23 01:06:43 -0500 received badge  Popular Question (source)
2016-09-22 01:46:43 -0500 asked a question Magnum create bay failed

Hi I am using openstack mitaka, and have followed the official doc to install magnum (version 2.0.0). After creating baymodel successfully, I got the error when create the bay.

Here is the debug information

Here is the baymodel

And my network

Does anyone get the same error as me?

Thanks a lot

2016-09-21 08:26:52 -0500 asked a question Magnum create bay failed Failed to create trustee or trust for Cluster


I am now using openstack mitaka and have installed magnum. The version of magnum is 2.0.0. After created bay-model successfully, I got an error while creating kubernetes bay. The error showed "Failed to create trustee or trust for Cluster".

Here's the log message (Imgur)

My /etc/magnum/magnum.conf


memcached_servers = controller:11211

auth_version = v3

auth_uri = http://controller:5000/v3

project_domain_name = default

project_name = service

user_domain_name = default

password = XXXXXX

username = magnum

auth_url = http://controller:35357

auth_type = password


trustee_domain_name = magnum

trustee_domain_admin_name = magnum_domain_admin

trustee_domain_admin_password = XXXXXX

I followed the official doc (

I have read this page and checked again and again that I didn't have the mistake on the setting. ( p.s. I can't run the script this page suggests cause the error below

"TypeError: __init__() takes at least 2 arguments (1 given) "

Does someone have this problem before and solved it?

Thanks a lot!

2016-09-18 23:04:25 -0500 commented answer Magnum commands return ERROR: Not Authorized

Thanks a lot!!! I got the same error as you. The document has too many errors...

2016-09-15 01:51:01 -0500 received badge  Enthusiast
2016-09-10 11:17:36 -0500 answered a question Connection to glance host http://controller:9292 failed

I solved my problem! I got the error because the firewall made rabbitmq.service failed. Try closed your firewall and restart your rabbitmq.service.

2016-09-07 13:59:28 -0500 answered a question Mitaka Nova Create VM failed caused by get glance image failed

Thank you for your response! I have solved the problem today, It's not a conspicuous error.

First of all, I can ping my controller from compute node successfully, and all the settings are right. After reading the nova-compute.log again and again, I found that neutron service get the similar error as glance. I started to think that the error was not in the glance but in basic settings. Finally, the error was in rabbitmq.service and firewall. It seems that firewall will make rabbitmq.service failed. It's interesting that though I closed the firewall, started the rabbitmq.service successfully in the beginning, restart the firewall, and checked the rabbitmq.service working well again. As firewall keep working, the rabbitmq.service failed even if its status was running. So, I closed my firewall, restart the rabbitmq.service, and didn't restart the firewall. The result was great, I successfully create the VM!

2016-09-07 08:08:32 -0500 received badge  Notable Question (source)
2016-09-06 07:21:15 -0500 received badge  Popular Question (source)
2016-09-05 21:07:18 -0500 commented question Connection to glance host http://controller:9292 failed

I don't have the same error in these two link. I can create the image using glance api successfully, and the controller:9292 is start.

2016-09-05 21:07:17 -0500 commented question Connection to glance host http://controller:9292 failed

Hey! I got the same error as you. Have you solved the problem?