nova ERROR:[Errno 111]连接被拒绝

时间:2014-03-24 17:54:41

标签: openstack

我正在使用Centos 6.5 x86_64来设置Openstack Havana并且所有服务都运行良好。但是当我重新启动操作系统时,我发现nova服务无法正常工作,触发了以下错误:

nova flavor-list
ERROR: [Errno 111] Connection refused

查看/ var / log / nova中的日志文件会出现以下错误:

2014-03-24 12:24:04.293 6275 INFO nova.osapi_compute.wsgi.server [-] (6275) wsgi starting up  
2014-03-24 12:24:04.297 6267 CRITICAL nova [-] [Errno 98] Address already in use 
2014-03-24 12:24:04.412 6275 INFO nova.openstack.common.service [-] Parent process has died unexpectedly, exiting 
2014-03-24 12:24:04.412 6274 INFO nova.openstack.common.service [-] Parent process has died unexpectedly, exiting 
2014-03-24 12:24:04.412 6275 INFO nova.wsgi [-] Stopping WSGI server. 
2014-03-24 12:24:04.412 6274 INFO nova.wsgi [-] Stopping WSGI server.

我的OpenStack服务器的状态

nova-manage service list

Binary           Host                                 Zone             Status     State Updated_At
nova-cert        controller                           internal         enabled    :-)   2014-03-24 14:28:03
nova-consoleauth controller                           internal         enabled    :-)   2014-03-24 14:28:01
nova-scheduler   controller                           internal         enabled    :-)   2014-03-24 14:28:00
nova-conductor   controller                           internal         enabled    :-)   2014-03-24 14:27:59
nova-compute     controller                           nova             enabled    :-)   2014-03-24 14:28:06
nova-network     controller                           internal         enabled    :-)   2014-03-24 14:27:58

keystone service-list

+----------------------------------+----------+----------+---------------------------+
|                id                |   name   |   type   |        description        |
+----------------------------------+----------+----------+---------------------------+
| 7ce108d652ee48d7897127045a371795 |  cinder  |  volume  |   Cinder Volume Service   |
| 9452b875328f4763b7766eb533bd75c4 | cinderv2 | volumev2 |  Cinder Volume Service V2 |
| e9607d1a308140298f8364fd2a0e62a8 |  glance  |  image   |    Glance Image Service   |
| b7ac07f69e2e41f684d6470c69db4781 | keystone | identity | Keystone Identity Service |
| cbdfa73329094d7d94c7464b9bf0ef7d |   nova   | compute  |    Nova Compute service   |
+----------------------------------+----------+----------+---------------------------+

ps -ef | grep "nova-api"

nova      2522     1  0 11:22 ?        00:00:00 /usr/bin/python /usr/bin/nova-api-metadata --logfile /var/log/nova/metadata-api.log
root     11909  6217  0 15:11 pts/1    00:00:01 gedit nova-api.log
root     12644  3832  0 15:31 pts/0    00:00:00 grep nova-api


netstat -napo | grep 877

tcp        0      0 0.0.0.0:8775                0.0.0.0:*                   LISTEN      2522/python         off (0.00/0/0)

任何指针都会非常有用。

由于

2 个答案:

答案 0 :(得分:2)

首先,我强烈建议您在ask.openstack.org上找到或要求回答

然后根据您的描述,它可能由以下原因引起:您已同时启用nova-api-metadata和nova-api服务。

从默认配置我们知道:['ec2','osapi_compute','元数据']已启用,请参阅https://github.com/openstack/nova/blob/stable/havana/nova/service.py#L55

所以当调用nova-api服务时它将逐个启动每个服务,请参阅https://github.com/openstack/nova/blob/stable/havana/nova/cmd/api.py#L45

由于nova-api-metadata服务正在运行,导致8775端口被使用,然后nova-api启动的一个服务将死亡,因为这个异常没有被捕获,那么其他两个也会死掉,然后你得到您在日志中看到的内容

如果我认为是对的,请取消nova-api-metadata服务并仅使用nova-api服务,这意味着'chkconfig openstack-nova-api-metadata off; chkconfig openstack-nova-api on',我不确定你系统上的具体服务名称,但是应该是这样的,如果我错了就纠正它

答案 1 :(得分:0)

拒绝连接是每次遇到的常见错误。其中一个案例是keystone拒绝新星服务的连接。

确保在创建keystone服务时,nova和quantum的SERVICE_PASSWORD相同。转到quantum和nova配置文件并验证SERVICE_PASSWORD是否相同。

恩乔伊!!