发起问题

Senol Erdogan 的档案 - activity

2019-07-27 00:06:10 -0500 评论问题 Install Question

For learn and understand the installing, the Installing Guideis going to be best place instead of using another resource for starting. docs.openstack.org has everything that you need to know. Installing guide needs to be read to save time and effort.

2019-07-26 23:46:00 -0500 回答问题 Deploy openstack in docker swarm

I've been testing keystone with multiple keystone server over swarm overlay network to see its swarm's load balancing . I didn't use another tool. Api, db, memcached and rabbitmq are in same stack as different 4 services. Only Api needs customized entrypoint. For other containers (db, memcached and rabbitmq) environment variables enough to start.

Glance's image, nova's guests disks and cinder's (if you are going to use LMV backend) disturbed block storage like Ceph for mobility, Or need to set these container's placement constraints in the swarm compose file to spawn them in the same host.

For neutron, heaven knows for now. It is going to be fun!

2019-07-26 23:17:02 -0500 回答问题 compute node looking for glance image over itself

Probably you forgot to put "api_servers = http://{glance_api_endpoint}:9292" in the nova.conf under [glance] section

2019-07-23 08:53:45 -0500 获得奖牌  复兴的问题 (source)
2019-07-22 15:29:06 -0500 回答问题 How to make username unique per project

Usernames are unique to domain-wide. Same username can be usable in another domain.

2019-07-22 15:20:16 -0500 回答问题 execute openstack service create xxxxx return http 500 error

The "mod_wsgi" and Python are not compatible. Probable the "mod_wsgi" is for Python2.7 and your setup needs the module for Python3x.

If you use Python3x, it needs to be seen like this

[root@node]# ls -al /usr/lib/apache2/modules | grep wsgi
lrwxrwxrwx. 1 root root 15 Feb 2 23:30 mod_wsgi.so -> mod_wsgi.so-3.7
-rw-r--r--. 1 root root 245424 Feb 2 23:30 mod_wsgi.so-3.7

2017-10-31 01:54:26 -0500 回答问题 Deploy two cloud environment with a single external storage

Cinder backends use single database and message queue to keep disk metadata and provision in the cloud environment. Cinder disks are nothing without is metadata in terms of using in cloud.

2017-09-18 03:03:29 -0500 评论回答 instance build

Seems there is a confusion needs some corrections. On the Controller node "nova-api.log, nova-scheduler.log,nova-conductor.log " and on the Compute node, "nova-compute.log"

and, if the instance still remains in "BUILD" state, can you post the result of

nova show provider-instance
2017-09-17 13:56:22 -0500 获得奖牌  拥护者 (source)
2017-09-17 13:52:10 -0500 评论回答 instance build

What compute logs say? Before the creation command;

  • On the controller node; tail -f /var/log/nova/nova-api.log /var/log/nova/nova-conductor.log /var/log/nova/nova-scheduler.log |grep ERROR

  • On the compute nodes; tail -f /var/log/nova/nova-compute.log |grep ERROR

2017-09-15 01:30:38 -0500 评论问题 instance build

Hi Costas, Dashboard doesn't give too much fault information. You need to start checking log files. The link in your question explain how to create instance via CLI. If you use CLI, you can add "--debug" parameter in to the command to see what is going wrong while your command is running.

2017-09-01 00:50:33 -0500 获得奖牌  批评家 (source)
2017-08-26 02:29:33 -0500 评论问题 How do I resync openstack heat after manually deleting nova host?

You need to comment out the instance (prod2nodessdsf02) in template file and update the stack with new commented out template file (heat stack-update).

2017-08-26 02:10:34 -0500 回答问题 Default Flavors Extra Specs

You can set extra specs with openstack or nova CLI; "nova flavor-key" "openstack flavor set"

Flavors create-update-delete operations are allowed for admin role by default nova policy.

2017-08-26 01:47:09 -0500 回答问题 Logs location configuration

Did you check nova.conf for logging details. There is a "log_dir=/var/log/nova" line that you can check. (reboot nova services after changes)

OR

Another way, in nova.conf you can enable "use_syslog=true" and then redirect the nova services logs to LOCAL0 with "syslog_log_facility=LOG_LOCAL0" (reboot nova services). So, redirected logs will be received by rsyslog daemon after that setting. Received log can be sent to a file with "LOCAL0.* /var/log/nova/nova.log', or to another log server with "LOCAL0.* @@log_server_ip_address" (reboot rsyslog)

Here you are an official doc about logging

2017-08-26 01:12:20 -0500 评论问题 Floating IP Association vs Floating IP

If the port has one Fixed IP, your template's approach is ok. If the port has more than one Fixed IP, one of them has to be chosen for Floating IP's forwarded traffic. Selecting a Fixed IP is a feature of "OS::Neutron::FloatingIPAssociation"'s "fixed_ip_address" property.

2016-09-21 10:55:28 -0500 获得奖牌  优秀回答 (source)
2015-08-15 08:15:01 -0500 获得奖牌  粉丝
2015-08-12 03:02:37 -0500 获得奖牌  好回答 (source)
2015-08-11 20:21:59 -0500 回答问题 can not sync database Ubuntu 15.04

First, enable your SQL server's query loging. Then watch SQL logs while trying the sync command. If tjere is any mistake in keystone.conf then you will see if your connection credential right or not in SQL logs. If they are not same which in logs and conf file, probably "conmection" setting is at wrong place. Move sql connection information from [default] to [database] section.

Second, disable firewall or give permission where installed SQL server.

After solve the connection problem you can disable SQL query loging.

2014-12-08 05:50:58 -0500 获得奖牌  好回答 (source)
2014-11-17 14:25:34 -0500 获得奖牌  复兴的问题 (source)
2014-11-17 14:25:34 -0500 获得奖牌  老师 (source)
2014-10-16 10:55:52 -0500 回答问题 Neutron net-create requires that tenant_id is specified

command needs admin tenant-id

firstly, take tenant list

 *# keystone tenant-list*

command will return tenants list as you know. such as;

| id | name | enabled | +-----------------------------------------------+----------+-----------+ | 6df22af2a590453ba8488dad5204b138 | admin | True | | a7fef4f8426d49b6b79ff007eb0117a8 | demo | True | | c976cd00104b40e59b7dd7d360aba6e4 | service | True ...

then, get admin tenant-id from list and use in the "neutron net-create" command, like

 *#neutron net-create ext-net --shared --router:external=True --tenant-id 6df22af2a590453ba8488dad5204b138*