)]}'
{"doc/source/specs/across_multiple_openstack.rst":[{"author":{"_account_id":7805,"name":"Li Ma","email":"skywalker.nick@gmail.com","username":"Nick"},"change_message_id":"76a61f9e55a79731e49fec0fff6c5c3b422ddd1c","unresolved":false,"context_lines":[{"line_number":5,"context_line":" http://creativecommons.org/licenses/by/3.0/legalcode\r"},{"line_number":6,"context_line":"\r"},{"line_number":7,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\r"},{"line_number":8,"context_line":"Communication Across Multiple OpenStack\r"},{"line_number":9,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\r"},{"line_number":10,"context_line":"\r"},{"line_number":11,"context_line":"This spec proposes the solutions of Communication across multiple OpenStack\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_76db52ee","line":8,"range":{"start_line":8,"start_character":21,"end_line":8,"end_character":39},"updated":"2016-06-06 01:41:00.000000000","message":"multiple openstack  or multi-region deployment?","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":20287,"name":"hujie","email":"mike.hu@huawei.com","username":"hujie"},"change_message_id":"96a748deabf3526905781b5d588195a8ec6e3cb4","unresolved":false,"context_lines":[{"line_number":5,"context_line":" http://creativecommons.org/licenses/by/3.0/legalcode\r"},{"line_number":6,"context_line":"\r"},{"line_number":7,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\r"},{"line_number":8,"context_line":"Communication Across Multiple OpenStack\r"},{"line_number":9,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\r"},{"line_number":10,"context_line":"\r"},{"line_number":11,"context_line":"This spec proposes the solutions of Communication across multiple OpenStack\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_f8f9e54e","line":8,"range":{"start_line":8,"start_character":21,"end_line":8,"end_character":39},"in_reply_to":"7aa08908_76db52ee","updated":"2016-06-06 08:59:58.000000000","message":"multiple independent openstack system, each system is a whole openstack include nova, neutron, keystone, cinder.......","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":20297,"name":"duankebo","email":"duankebo@huawei.com","username":"duankebo"},"change_message_id":"7de3b5e7310655c2eb05497ce42a673f5e0f846a","unresolved":false,"context_lines":[{"line_number":16,"context_line":"\r"},{"line_number":17,"context_line":"In the large scale deployment situation, we will deploy a large number of\r"},{"line_number":18,"context_line":"compute nodes. However, the management ability of one OpenStack system is\r"},{"line_number":19,"context_line":"limited, when the number of compute nodes is large enough, the ability of\r"},{"line_number":20,"context_line":"DF DB and the performance of sub/pub process become a big problem.\r"},{"line_number":21,"context_line":"\r"},{"line_number":22,"context_line":"So we must divide the many compute nodes into different regions, each region\r"},{"line_number":23,"context_line":"is managed by one OpenStack which has the independent DF DB and sub/pub\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_87dd8a44","line":20,"range":{"start_line":19,"start_character":59,"end_line":20,"end_character":44},"updated":"2016-06-11 15:07:34.000000000","message":"you means performance of df db and sub/pub will be a big problem? I don\u0027t think it\u0027s the primary reason. In fact vms in a openstack need to interact with other entities such as physical machines, routers or firewalls that are not managed by dragonflow. In order to communicate with them, we need port info these devices.","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":7805,"name":"Li Ma","email":"skywalker.nick@gmail.com","username":"Nick"},"change_message_id":"76a61f9e55a79731e49fec0fff6c5c3b422ddd1c","unresolved":false,"context_lines":[{"line_number":19,"context_line":"limited, when the number of compute nodes is large enough, the ability of\r"},{"line_number":20,"context_line":"DF DB and the performance of sub/pub process become a big problem.\r"},{"line_number":21,"context_line":"\r"},{"line_number":22,"context_line":"So we must divide the many compute nodes into different regions, each region\r"},{"line_number":23,"context_line":"is managed by one OpenStack which has the independent DF DB and sub/pub\r"},{"line_number":24,"context_line":"system. We could introduce a higher level Orchestrator to manage the multiple\r"},{"line_number":25,"context_line":"OpenStack regions.\r"},{"line_number":26,"context_line":"\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_768992ca","line":23,"range":{"start_line":22,"start_character":46,"end_line":23,"end_character":27},"updated":"2016-06-06 01:41:00.000000000","message":"The same question? Each region has an independent neutron endpoint, which is registered in the shared keystone instance. I cannot get it.","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":20287,"name":"hujie","email":"mike.hu@huawei.com","username":"hujie"},"change_message_id":"96a748deabf3526905781b5d588195a8ec6e3cb4","unresolved":false,"context_lines":[{"line_number":19,"context_line":"limited, when the number of compute nodes is large enough, the ability of\r"},{"line_number":20,"context_line":"DF DB and the performance of sub/pub process become a big problem.\r"},{"line_number":21,"context_line":"\r"},{"line_number":22,"context_line":"So we must divide the many compute nodes into different regions, each region\r"},{"line_number":23,"context_line":"is managed by one OpenStack which has the independent DF DB and sub/pub\r"},{"line_number":24,"context_line":"system. We could introduce a higher level Orchestrator to manage the multiple\r"},{"line_number":25,"context_line":"OpenStack regions.\r"},{"line_number":26,"context_line":"\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_d85ea944","line":23,"range":{"start_line":22,"start_character":46,"end_line":23,"end_character":27},"in_reply_to":"7aa08908_768992ca","updated":"2016-06-06 08:59:58.000000000","message":"each openstack is a independent system include separate keystone and neutron.","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":7805,"name":"Li Ma","email":"skywalker.nick@gmail.com","username":"Nick"},"change_message_id":"76a61f9e55a79731e49fec0fff6c5c3b422ddd1c","unresolved":false,"context_lines":[{"line_number":25,"context_line":"OpenStack regions.\r"},{"line_number":26,"context_line":"\r"},{"line_number":27,"context_line":"If two virtual machines which belong to the same network are divided into\r"},{"line_number":28,"context_line":"different OpenStack Region, we must resolve the communication problem between\r"},{"line_number":29,"context_line":"the two virtual machines.\r"},{"line_number":30,"context_line":"\r"},{"line_number":31,"context_line":"Proposed Change\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_1672cee5","line":28,"range":{"start_line":28,"start_character":48,"end_line":28,"end_character":69},"updated":"2016-06-06 01:41:00.000000000","message":"In the multi-region deployment, we always use floating ips to communicate virtual machines across regions. Could you elaborate it why this approach is better.","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":7805,"name":"Li Ma","email":"skywalker.nick@gmail.com","username":"Nick"},"change_message_id":"76a61f9e55a79731e49fec0fff6c5c3b422ddd1c","unresolved":false,"context_lines":[{"line_number":24,"context_line":"system. We could introduce a higher level Orchestrator to manage the multiple\r"},{"line_number":25,"context_line":"OpenStack regions.\r"},{"line_number":26,"context_line":"\r"},{"line_number":27,"context_line":"If two virtual machines which belong to the same network are divided into\r"},{"line_number":28,"context_line":"different OpenStack Region, we must resolve the communication problem between\r"},{"line_number":29,"context_line":"the two virtual machines.\r"},{"line_number":30,"context_line":"\r"},{"line_number":31,"context_line":"Proposed Change\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_96a7de4e","line":28,"range":{"start_line":27,"start_character":40,"end_line":28,"end_character":26},"updated":"2016-06-06 01:41:00.000000000","message":"What you wanna deal with is to create a \u0027SAME\u0027 network for multiple neutron instances across regions and share the same attributes and also in the single control plane? Am I right? Is this spec related with tricircle or your cascading solution?","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":20287,"name":"hujie","email":"mike.hu@huawei.com","username":"hujie"},"change_message_id":"96a748deabf3526905781b5d588195a8ec6e3cb4","unresolved":false,"context_lines":[{"line_number":24,"context_line":"system. We could introduce a higher level Orchestrator to manage the multiple\r"},{"line_number":25,"context_line":"OpenStack regions.\r"},{"line_number":26,"context_line":"\r"},{"line_number":27,"context_line":"If two virtual machines which belong to the same network are divided into\r"},{"line_number":28,"context_line":"different OpenStack Region, we must resolve the communication problem between\r"},{"line_number":29,"context_line":"the two virtual machines.\r"},{"line_number":30,"context_line":"\r"},{"line_number":31,"context_line":"Proposed Change\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_9bca5ba0","line":28,"range":{"start_line":27,"start_character":40,"end_line":28,"end_character":26},"in_reply_to":"7aa08908_96a7de4e","updated":"2016-06-06 08:59:58.000000000","message":"yes, it\u0027s a cascading solution, and a high level Openstack system which we call it Orchestrator could manage many independent openstack systems, and it will create a \u0027same\u0027 network for the openstack systems it manage, and the communication between two vms from dif openstack region is using the normal tunnel encap, not floating ips, because we do not across Internet","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":7805,"name":"Li Ma","email":"skywalker.nick@gmail.com","username":"Nick"},"change_message_id":"76a61f9e55a79731e49fec0fff6c5c3b422ddd1c","unresolved":false,"context_lines":[{"line_number":31,"context_line":"Proposed Change\r"},{"line_number":32,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\r"},{"line_number":33,"context_line":"\r"},{"line_number":34,"context_line":"To resolve the problem, the general idea is Neutron plugin analyses the\r"},{"line_number":35,"context_line":"notification message which indicates a remote Neutron port has been created\r"},{"line_number":36,"context_line":"sent by the Orchestrator layer and finds it\u0027s a remote Neutron port which\r"},{"line_number":37,"context_line":"comes from another OpenStack region, plugin will assign a specific chassis\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_967ebeac","line":34,"range":{"start_line":34,"start_character":59,"end_line":34,"end_character":67},"updated":"2016-06-06 01:41:00.000000000","message":"analyzes","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":7805,"name":"Li Ma","email":"skywalker.nick@gmail.com","username":"Nick"},"change_message_id":"76a61f9e55a79731e49fec0fff6c5c3b422ddd1c","unresolved":false,"context_lines":[{"line_number":32,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\r"},{"line_number":33,"context_line":"\r"},{"line_number":34,"context_line":"To resolve the problem, the general idea is Neutron plugin analyses the\r"},{"line_number":35,"context_line":"notification message which indicates a remote Neutron port has been created\r"},{"line_number":36,"context_line":"sent by the Orchestrator layer and finds it\u0027s a remote Neutron port which\r"},{"line_number":37,"context_line":"comes from another OpenStack region, plugin will assign a specific chassis\r"},{"line_number":38,"context_line":"name for the port and publish the message. The corresponding chassis will\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_b6581a58","line":35,"range":{"start_line":35,"start_character":39,"end_line":35,"end_character":59},"updated":"2016-06-06 01:41:00.000000000","message":"What is remote neutron port? Is it in the current neutron implementation? Is there any binding:profile associated with this remote type?","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":20287,"name":"hujie","email":"mike.hu@huawei.com","username":"hujie"},"change_message_id":"96a748deabf3526905781b5d588195a8ec6e3cb4","unresolved":false,"context_lines":[{"line_number":32,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\r"},{"line_number":33,"context_line":"\r"},{"line_number":34,"context_line":"To resolve the problem, the general idea is Neutron plugin analyses the\r"},{"line_number":35,"context_line":"notification message which indicates a remote Neutron port has been created\r"},{"line_number":36,"context_line":"sent by the Orchestrator layer and finds it\u0027s a remote Neutron port which\r"},{"line_number":37,"context_line":"comes from another OpenStack region, plugin will assign a specific chassis\r"},{"line_number":38,"context_line":"name for the port and publish the message. The corresponding chassis will\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_bb6f5791","line":35,"range":{"start_line":35,"start_character":39,"end_line":35,"end_character":59},"in_reply_to":"7aa08908_b6581a58","updated":"2016-06-06 08:59:58.000000000","message":"the remote neutron port here means the remote port comes from another independent openstack system region, we could consider add the info which indicate it is a remote port into binding_profile field","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":7805,"name":"Li Ma","email":"skywalker.nick@gmail.com","username":"Nick"},"change_message_id":"76a61f9e55a79731e49fec0fff6c5c3b422ddd1c","unresolved":false,"context_lines":[{"line_number":33,"context_line":"\r"},{"line_number":34,"context_line":"To resolve the problem, the general idea is Neutron plugin analyses the\r"},{"line_number":35,"context_line":"notification message which indicates a remote Neutron port has been created\r"},{"line_number":36,"context_line":"sent by the Orchestrator layer and finds it\u0027s a remote Neutron port which\r"},{"line_number":37,"context_line":"comes from another OpenStack region, plugin will assign a specific chassis\r"},{"line_number":38,"context_line":"name for the port and publish the message. The corresponding chassis will\r"},{"line_number":39,"context_line":"receive the message because of its previously subscribe, and it will create\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_363b8aed","line":36,"range":{"start_line":36,"start_character":12,"end_line":36,"end_character":24},"updated":"2016-06-06 01:41:00.000000000","message":"What is the orchestrator? I think you should provide a section to describe it.","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":20287,"name":"hujie","email":"mike.hu@huawei.com","username":"hujie"},"change_message_id":"96a748deabf3526905781b5d588195a8ec6e3cb4","unresolved":false,"context_lines":[{"line_number":33,"context_line":"\r"},{"line_number":34,"context_line":"To resolve the problem, the general idea is Neutron plugin analyses the\r"},{"line_number":35,"context_line":"notification message which indicates a remote Neutron port has been created\r"},{"line_number":36,"context_line":"sent by the Orchestrator layer and finds it\u0027s a remote Neutron port which\r"},{"line_number":37,"context_line":"comes from another OpenStack region, plugin will assign a specific chassis\r"},{"line_number":38,"context_line":"name for the port and publish the message. The corresponding chassis will\r"},{"line_number":39,"context_line":"receive the message because of its previously subscribe, and it will create\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_7b2b2f31","line":36,"range":{"start_line":36,"start_character":12,"end_line":36,"end_character":24},"in_reply_to":"7aa08908_363b8aed","updated":"2016-06-06 08:59:58.000000000","message":"It\u0027s a high level Openstack system","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":20297,"name":"duankebo","email":"duankebo@huawei.com","username":"duankebo"},"change_message_id":"7de3b5e7310655c2eb05497ce42a673f5e0f846a","unresolved":false,"context_lines":[{"line_number":46,"context_line":"When a remote Neutron port which belongs to a VM is created in remote\r"},{"line_number":47,"context_line":"OpenStack region, the Orchestrator layer will send a create_port notification\r"},{"line_number":48,"context_line":"message to local OpenStack region, and Local Neutron plugin will process it,\r"},{"line_number":49,"context_line":"the details show below:\r"},{"line_number":50,"context_line":"\r"},{"line_number":51,"context_line":"1. We could assume the binding_profile field contains the info that indicate\r"},{"line_number":52,"context_line":"the Neutron port is a remote port comes from another OpenStack region, the\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_67b7bef5","line":49,"range":{"start_line":49,"start_character":12,"end_line":49,"end_character":16},"updated":"2016-06-11 15:07:34.000000000","message":"are shown","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":7805,"name":"Li Ma","email":"skywalker.nick@gmail.com","username":"Nick"},"change_message_id":"76a61f9e55a79731e49fec0fff6c5c3b422ddd1c","unresolved":false,"context_lines":[{"line_number":52,"context_line":"the Neutron port is a remote port comes from another OpenStack region, the\r"},{"line_number":53,"context_line":"field could be like this:\r"},{"line_number":54,"context_line":"\r"},{"line_number":55,"context_line":"binding_profile \u003d {\"port_key\": \"remote_port\",\r"},{"line_number":56,"context_line":"                   \"host_ip\": remote_chassis_ip}\r"},{"line_number":57,"context_line":"\r"},{"line_number":58,"context_line":"2. When Neutron plugin find it is a remote port by the binding_profile field\r"},{"line_number":59,"context_line":"in the create_port message, it will generate a specific chassis name for the\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_d103300c","line":56,"range":{"start_line":55,"start_character":0,"end_line":56,"end_character":48},"updated":"2016-06-06 01:41:00.000000000","message":"OK, so it is a kind of binding-profile attribute which I guess. Which open source neutron backend does support it? I doubt.","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":20287,"name":"hujie","email":"mike.hu@huawei.com","username":"hujie"},"change_message_id":"5a2a3ac086dce60feb0f542d63f58a3b89a5d07c","unresolved":false,"context_lines":[{"line_number":52,"context_line":"the Neutron port is a remote port comes from another OpenStack region, the\r"},{"line_number":53,"context_line":"field could be like this:\r"},{"line_number":54,"context_line":"\r"},{"line_number":55,"context_line":"binding_profile \u003d {\"port_key\": \"remote_port\",\r"},{"line_number":56,"context_line":"                   \"host_ip\": remote_chassis_ip}\r"},{"line_number":57,"context_line":"\r"},{"line_number":58,"context_line":"2. When Neutron plugin find it is a remote port by the binding_profile field\r"},{"line_number":59,"context_line":"in the create_port message, it will generate a specific chassis name for the\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_e1727e7b","line":56,"range":{"start_line":55,"start_character":0,"end_line":56,"end_character":48},"in_reply_to":"7aa08908_a7357447","updated":"2016-06-11 10:03:58.000000000","message":"Yes, I know what you mean, I\u0027ll consider how to deal with this patch:)","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":20287,"name":"hujie","email":"mike.hu@huawei.com","username":"hujie"},"change_message_id":"96a748deabf3526905781b5d588195a8ec6e3cb4","unresolved":false,"context_lines":[{"line_number":52,"context_line":"the Neutron port is a remote port comes from another OpenStack region, the\r"},{"line_number":53,"context_line":"field could be like this:\r"},{"line_number":54,"context_line":"\r"},{"line_number":55,"context_line":"binding_profile \u003d {\"port_key\": \"remote_port\",\r"},{"line_number":56,"context_line":"                   \"host_ip\": remote_chassis_ip}\r"},{"line_number":57,"context_line":"\r"},{"line_number":58,"context_line":"2. When Neutron plugin find it is a remote port by the binding_profile field\r"},{"line_number":59,"context_line":"in the create_port message, it will generate a specific chassis name for the\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_dba3439d","line":56,"range":{"start_line":55,"start_character":0,"end_line":56,"end_character":48},"in_reply_to":"7aa08908_d103300c","updated":"2016-06-06 08:59:58.000000000","message":"the Orchestrator could add the info into this field to indicate it\u0027s a remote port, you can say it\u0027s a temporary method, but I think binding_profile field is suitable to catch the info if we want to notify it\u0027s a remote port","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":7805,"name":"Li Ma","email":"skywalker.nick@gmail.com","username":"Nick"},"change_message_id":"97edc59990319903589f2208e754321d05ce70d0","unresolved":false,"context_lines":[{"line_number":52,"context_line":"the Neutron port is a remote port comes from another OpenStack region, the\r"},{"line_number":53,"context_line":"field could be like this:\r"},{"line_number":54,"context_line":"\r"},{"line_number":55,"context_line":"binding_profile \u003d {\"port_key\": \"remote_port\",\r"},{"line_number":56,"context_line":"                   \"host_ip\": remote_chassis_ip}\r"},{"line_number":57,"context_line":"\r"},{"line_number":58,"context_line":"2. When Neutron plugin find it is a remote port by the binding_profile field\r"},{"line_number":59,"context_line":"in the create_port message, it will generate a specific chassis name for the\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_a7357447","line":56,"range":{"start_line":55,"start_character":0,"end_line":56,"end_character":48},"in_reply_to":"7aa08908_dba3439d","updated":"2016-06-06 10:14:12.000000000","message":"I wonder if the mechanism is available in tricircle. How to test it when you have done the work in dragonflow is very important to demonstrate it is working and is not able to affect other deployments. Currently there is no neutron backend that supports remote port, even ML2 ovs/linuxbridge, as far as I know.","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":20297,"name":"duankebo","email":"duankebo@huawei.com","username":"duankebo"},"change_message_id":"7de3b5e7310655c2eb05497ce42a673f5e0f846a","unresolved":false,"context_lines":[{"line_number":66,"context_line":"-------------------\r"},{"line_number":67,"context_line":"\r"},{"line_number":68,"context_line":"The corresponding chassis will receive the create_port message published by\r"},{"line_number":69,"context_line":"Neutron plugin because of its previously subscribe:\r"},{"line_number":70,"context_line":"\r"},{"line_number":71,"context_line":"1. DF local controller will analyse the create_port message and find it is a\r"},{"line_number":72,"context_line":"remote OpenStack region port by the specific chassis name, and also it will\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_27a496b7","line":69,"range":{"start_line":69,"start_character":30,"end_line":69,"end_character":40},"updated":"2016-06-11 15:07:34.000000000","message":"When do a chassis subscribe message of this kind ?","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":20297,"name":"duankebo","email":"duankebo@huawei.com","username":"duankebo"},"change_message_id":"7de3b5e7310655c2eb05497ce42a673f5e0f846a","unresolved":false,"context_lines":[{"line_number":66,"context_line":"-------------------\r"},{"line_number":67,"context_line":"\r"},{"line_number":68,"context_line":"The corresponding chassis will receive the create_port message published by\r"},{"line_number":69,"context_line":"Neutron plugin because of its previously subscribe:\r"},{"line_number":70,"context_line":"\r"},{"line_number":71,"context_line":"1. DF local controller will analyse the create_port message and find it is a\r"},{"line_number":72,"context_line":"remote OpenStack region port by the specific chassis name, and also it will\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_47a1e2a8","line":69,"range":{"start_line":69,"start_character":41,"end_line":69,"end_character":50},"updated":"2016-06-11 15:07:34.000000000","message":"subscription","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":20297,"name":"duankebo","email":"duankebo@huawei.com","username":"duankebo"},"change_message_id":"7de3b5e7310655c2eb05497ce42a673f5e0f846a","unresolved":false,"context_lines":[{"line_number":74,"context_line":"\r"},{"line_number":75,"context_line":"2. Local controller will check whether the chassis has the tunnel port from\r"},{"line_number":76,"context_line":"itself to the specific remote chassis, if not, it will create the tunnel\r"},{"line_number":77,"context_line":"port and establish the tunnel to the remote chassis.\r"},{"line_number":78,"context_line":"\r"},{"line_number":79,"context_line":"3. After the tunnel port has been created, local controller will notify the\r"},{"line_number":80,"context_line":"create_lport message to Apps, it will be considered a normal remote port as\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_c795d240","line":77,"range":{"start_line":77,"start_character":23,"end_line":77,"end_character":36},"updated":"2016-06-11 15:07:34.000000000","message":"will this lead to too many tunnel are set up?","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"},{"author":{"_account_id":20297,"name":"duankebo","email":"duankebo@huawei.com","username":"duankebo"},"change_message_id":"7de3b5e7310655c2eb05497ce42a673f5e0f846a","unresolved":false,"context_lines":[{"line_number":81,"context_line":"in current implementation.\r"},{"line_number":82,"context_line":"\r"},{"line_number":83,"context_line":"On the other hand, when the last remote OpenStack region port on a remote\r"},{"line_number":84,"context_line":"chassis would be deleted from local cache, it means there are no need to\r"},{"line_number":85,"context_line":"communicate to the remote chassis anymore for the local controller, it should\r"},{"line_number":86,"context_line":"delete the corresponding tunnel port.\r"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7aa08908_4788c222","line":84,"range":{"start_line":84,"start_character":8,"end_line":84,"end_character":16},"updated":"2016-06-11 15:07:34.000000000","message":"is","commit_id":"da42928a8757132ec91578c975bcd45f11da51b1"}]}