)]}'
{"devstack/override-defaults":[{"author":{"_account_id":1561,"name":"Russell Bryant","email":"rbryant@redhat.com","username":"russellb"},"change_message_id":"7885d12f03f826c9c9370a5bb8b5f2cf2db7ef66","unresolved":false,"context_lines":[{"line_number":1,"context_line":"Q_PLUGIN\u003d${Q_PLUGIN:-\"ml2\"}"},{"line_number":2,"context_line":"Q_ML2_PLUGIN_MECHANISM_DRIVERS\u003d${Q_ML2_PLUGIN_MECHANISM_DRIVERS:-ovn,logger}"},{"line_number":3,"context_line":"Q_ML2_TENANT_NETWORK_TYPE\u003d${Q_ML2_TENANT_NETWORK_TYPE:-\"geneve\"}"},{"line_number":4,"context_line":"Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS\u003d${Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS:-\"vni_ranges\u003d1:4095\"}"},{"line_number":5,"context_line":""},{"line_number":6,"context_line":"# Whether to enable using OVN\u0027s L3 functionality. If this value is disabled,"},{"line_number":7,"context_line":"# OpenStack will use q-l3 functionality."}],"source_content_type":"application/octet-stream","patch_set":3,"id":"3aaa91ec_5b12ba63","line":4,"updated":"2016-06-22 01:39:39.000000000","message":"any reason not to make this a much larger range?  We default to a limit of 4095 networks?\n\nIn practice for a trivial devstack env, that\u0027s fine, I\u0027m just worried about how broadly this line is going to get copied around.","commit_id":"8162d9d622d851f58db20449a5d9e62d54c2ea4c"},{"author":{"_account_id":105,"name":"Kyle Mestery","email":"mestery@mestery.com","username":"mestery"},"change_message_id":"7df6d07416ec02f86db8c3ecfd87f1ed766afffd","unresolved":false,"context_lines":[{"line_number":1,"context_line":"Q_PLUGIN\u003d${Q_PLUGIN:-\"ml2\"}"},{"line_number":2,"context_line":"Q_ML2_PLUGIN_MECHANISM_DRIVERS\u003d${Q_ML2_PLUGIN_MECHANISM_DRIVERS:-ovn,logger}"},{"line_number":3,"context_line":"Q_ML2_TENANT_NETWORK_TYPE\u003d${Q_ML2_TENANT_NETWORK_TYPE:-\"geneve\"}"},{"line_number":4,"context_line":"Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS\u003d${Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS:-\"vni_ranges\u003d1:4095\"}"},{"line_number":5,"context_line":""},{"line_number":6,"context_line":"# Whether to enable using OVN\u0027s L3 functionality. If this value is disabled,"},{"line_number":7,"context_line":"# OpenStack will use q-l3 functionality."}],"source_content_type":"application/octet-stream","patch_set":3,"id":"3aaa91ec_fe64344a","line":4,"in_reply_to":"3aaa91ec_5b12ba63","updated":"2016-06-22 01:54:11.000000000","message":"I can see that as well, and I think we should bump this up as Russell suggests.","commit_id":"8162d9d622d851f58db20449a5d9e62d54c2ea4c"},{"author":{"_account_id":8410,"name":"Richard Theis","email":"rtheis@us.ibm.com","username":"rtheis"},"change_message_id":"a2208f5cdb7a4f1ce39844a9e04e450bb9eb394a","unresolved":false,"context_lines":[{"line_number":1,"context_line":"Q_PLUGIN\u003d${Q_PLUGIN:-\"ml2\"}"},{"line_number":2,"context_line":"Q_ML2_PLUGIN_MECHANISM_DRIVERS\u003d${Q_ML2_PLUGIN_MECHANISM_DRIVERS:-ovn,logger}"},{"line_number":3,"context_line":"Q_ML2_TENANT_NETWORK_TYPE\u003d${Q_ML2_TENANT_NETWORK_TYPE:-\"geneve\"}"},{"line_number":4,"context_line":"Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS\u003d${Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS:-\"vni_ranges\u003d1:4095\"}"},{"line_number":5,"context_line":""},{"line_number":6,"context_line":"# Whether to enable using OVN\u0027s L3 functionality. If this value is disabled,"},{"line_number":7,"context_line":"# OpenStack will use q-l3 functionality."}],"source_content_type":"application/octet-stream","patch_set":3,"id":"3aaa91ec_d8a04c93","line":4,"in_reply_to":"3aaa91ec_fe64344a","updated":"2016-06-22 13:53:53.000000000","message":"I originally tried 16777215 networks (which is the maximum) and that took too long to start neutron-server due to the number of segments that had to be created in the neutron DB.  I\u0027ll update with a more reasonable range.","commit_id":"8162d9d622d851f58db20449a5d9e62d54c2ea4c"},{"author":{"_account_id":105,"name":"Kyle Mestery","email":"mestery@mestery.com","username":"mestery"},"change_message_id":"b45d28a0239ae3b7be0f561a55f89a3459fa65aa","unresolved":false,"context_lines":[{"line_number":1,"context_line":"Q_PLUGIN\u003d${Q_PLUGIN:-\"ml2\"}"},{"line_number":2,"context_line":"Q_ML2_PLUGIN_MECHANISM_DRIVERS\u003d${Q_ML2_PLUGIN_MECHANISM_DRIVERS:-ovn,logger}"},{"line_number":3,"context_line":"Q_ML2_TENANT_NETWORK_TYPE\u003d${Q_ML2_TENANT_NETWORK_TYPE:-\"geneve\"}"},{"line_number":4,"context_line":"Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS\u003d${Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS:-\"vni_ranges\u003d1:65536\"}"},{"line_number":5,"context_line":""},{"line_number":6,"context_line":"# Whether to enable using OVN\u0027s L3 functionality. If this value is disabled,"},{"line_number":7,"context_line":"# OpenStack will use q-l3 functionality."}],"source_content_type":"application/octet-stream","patch_set":4,"id":"3aaa91ec_6a0dfb3d","line":4,"updated":"2016-06-22 15:14:01.000000000","message":"This seems like a reasonable range to me. Can you confirm this does not affect Neutron server startup as well? Thanks!","commit_id":"c6461e61ab5c3de96c2f2421b3794d1689c6194b"},{"author":{"_account_id":4395,"name":"Aaron Rosen","email":"aaronorosen@gmail.com","username":"arosen"},"change_message_id":"63c043273e1a7aae8b749f91a1398dc8a8777e61","unresolved":false,"context_lines":[{"line_number":1,"context_line":"Q_PLUGIN\u003d${Q_PLUGIN:-\"ml2\"}"},{"line_number":2,"context_line":"Q_ML2_PLUGIN_MECHANISM_DRIVERS\u003d${Q_ML2_PLUGIN_MECHANISM_DRIVERS:-ovn,logger}"},{"line_number":3,"context_line":"Q_ML2_TENANT_NETWORK_TYPE\u003d${Q_ML2_TENANT_NETWORK_TYPE:-\"geneve\"}"},{"line_number":4,"context_line":"Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS\u003d${Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS:-\"vni_ranges\u003d1:65536\"}"},{"line_number":5,"context_line":""},{"line_number":6,"context_line":"# Whether to enable using OVN\u0027s L3 functionality. If this value is disabled,"},{"line_number":7,"context_line":"# OpenStack will use q-l3 functionality."}],"source_content_type":"application/octet-stream","patch_set":4,"id":"3aaa91ec_70e676b5","line":4,"in_reply_to":"3aaa91ec_05ac1634","updated":"2016-06-22 21:55:24.000000000","message":"also i\u0027m guessing that ml2 already would have that problem anyways so we can fix that later if so.","commit_id":"c6461e61ab5c3de96c2f2421b3794d1689c6194b"},{"author":{"_account_id":8410,"name":"Richard Theis","email":"rtheis@us.ibm.com","username":"rtheis"},"change_message_id":"f02248a83ae408ab8223019c5792d266fda18fd3","unresolved":false,"context_lines":[{"line_number":1,"context_line":"Q_PLUGIN\u003d${Q_PLUGIN:-\"ml2\"}"},{"line_number":2,"context_line":"Q_ML2_PLUGIN_MECHANISM_DRIVERS\u003d${Q_ML2_PLUGIN_MECHANISM_DRIVERS:-ovn,logger}"},{"line_number":3,"context_line":"Q_ML2_TENANT_NETWORK_TYPE\u003d${Q_ML2_TENANT_NETWORK_TYPE:-\"geneve\"}"},{"line_number":4,"context_line":"Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS\u003d${Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS:-\"vni_ranges\u003d1:65536\"}"},{"line_number":5,"context_line":""},{"line_number":6,"context_line":"# Whether to enable using OVN\u0027s L3 functionality. If this value is disabled,"},{"line_number":7,"context_line":"# OpenStack will use q-l3 functionality."}],"source_content_type":"application/octet-stream","patch_set":4,"id":"3aaa91ec_ade89d11","line":4,"in_reply_to":"3aaa91ec_6a0dfb3d","updated":"2016-06-22 15:24:45.000000000","message":"I deployed DevStack with 4095 and then updated to 65536 after deployment.  The neutron server restarted very quickly with 65536 (about 15 seconds).  This is similar to what I saw with no change to vni_ranges.","commit_id":"c6461e61ab5c3de96c2f2421b3794d1689c6194b"},{"author":{"_account_id":8410,"name":"Richard Theis","email":"rtheis@us.ibm.com","username":"rtheis"},"change_message_id":"e13d39764f0bf092290daed9889cee5e9e29becf","unresolved":false,"context_lines":[{"line_number":1,"context_line":"Q_PLUGIN\u003d${Q_PLUGIN:-\"ml2\"}"},{"line_number":2,"context_line":"Q_ML2_PLUGIN_MECHANISM_DRIVERS\u003d${Q_ML2_PLUGIN_MECHANISM_DRIVERS:-ovn,logger}"},{"line_number":3,"context_line":"Q_ML2_TENANT_NETWORK_TYPE\u003d${Q_ML2_TENANT_NETWORK_TYPE:-\"geneve\"}"},{"line_number":4,"context_line":"Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS\u003d${Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS:-\"vni_ranges\u003d1:65536\"}"},{"line_number":5,"context_line":""},{"line_number":6,"context_line":"# Whether to enable using OVN\u0027s L3 functionality. If this value is disabled,"},{"line_number":7,"context_line":"# OpenStack will use q-l3 functionality."}],"source_content_type":"application/octet-stream","patch_set":4,"id":"3aaa91ec_e119db7a","line":4,"in_reply_to":"3aaa91ec_70e676b5","updated":"2016-06-23 12:51:50.000000000","message":"I think the ML2 code will handle such situations today.  It does a locked query against all with retry support. I believe this should prevent the concurrency issues with multiple neutron servers initializing the database.","commit_id":"c6461e61ab5c3de96c2f2421b3794d1689c6194b"},{"author":{"_account_id":8410,"name":"Richard Theis","email":"rtheis@us.ibm.com","username":"rtheis"},"change_message_id":"ef96dbf4a7955d7cabca3917f3ffe75d3c250ef2","unresolved":false,"context_lines":[{"line_number":1,"context_line":"Q_PLUGIN\u003d${Q_PLUGIN:-\"ml2\"}"},{"line_number":2,"context_line":"Q_ML2_PLUGIN_MECHANISM_DRIVERS\u003d${Q_ML2_PLUGIN_MECHANISM_DRIVERS:-ovn,logger}"},{"line_number":3,"context_line":"Q_ML2_TENANT_NETWORK_TYPE\u003d${Q_ML2_TENANT_NETWORK_TYPE:-\"geneve\"}"},{"line_number":4,"context_line":"Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS\u003d${Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS:-\"vni_ranges\u003d1:65536\"}"},{"line_number":5,"context_line":""},{"line_number":6,"context_line":"# Whether to enable using OVN\u0027s L3 functionality. If this value is disabled,"},{"line_number":7,"context_line":"# OpenStack will use q-l3 functionality."}],"source_content_type":"application/octet-stream","patch_set":4,"id":"3aaa91ec_b9e8db11","line":4,"in_reply_to":"3aaa91ec_ade89d11","updated":"2016-06-22 16:19:46.000000000","message":"DevStack deployment with 65536 took 35 seconds longer than one with 4095.  But there\u0027s a lot of variability in those numbers.","commit_id":"c6461e61ab5c3de96c2f2421b3794d1689c6194b"},{"author":{"_account_id":4395,"name":"Aaron Rosen","email":"aaronorosen@gmail.com","username":"arosen"},"change_message_id":"63c6ae3c85304eaf3ba5e7a5e566651f8d1ad867","unresolved":false,"context_lines":[{"line_number":1,"context_line":"Q_PLUGIN\u003d${Q_PLUGIN:-\"ml2\"}"},{"line_number":2,"context_line":"Q_ML2_PLUGIN_MECHANISM_DRIVERS\u003d${Q_ML2_PLUGIN_MECHANISM_DRIVERS:-ovn,logger}"},{"line_number":3,"context_line":"Q_ML2_TENANT_NETWORK_TYPE\u003d${Q_ML2_TENANT_NETWORK_TYPE:-\"geneve\"}"},{"line_number":4,"context_line":"Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS\u003d${Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS:-\"vni_ranges\u003d1:65536\"}"},{"line_number":5,"context_line":""},{"line_number":6,"context_line":"# Whether to enable using OVN\u0027s L3 functionality. If this value is disabled,"},{"line_number":7,"context_line":"# OpenStack will use q-l3 functionality."}],"source_content_type":"application/octet-stream","patch_set":4,"id":"3aaa91ec_05ac1634","line":4,"in_reply_to":"3aaa91ec_b9e8db11","updated":"2016-06-22 21:54:39.000000000","message":"i wonder if these is any issue if you start up multiple neutron servers at once. I guess the underlying code pre-populates entries into the db?","commit_id":"c6461e61ab5c3de96c2f2421b3794d1689c6194b"},{"author":{"_account_id":4395,"name":"Aaron Rosen","email":"aaronorosen@gmail.com","username":"arosen"},"change_message_id":"9b198d4fb9dbce916f09bfe48bac57c4ff92b428","unresolved":false,"context_lines":[{"line_number":1,"context_line":"Q_PLUGIN\u003d${Q_PLUGIN:-\"ml2\"}"},{"line_number":2,"context_line":"Q_ML2_PLUGIN_MECHANISM_DRIVERS\u003d${Q_ML2_PLUGIN_MECHANISM_DRIVERS:-ovn,logger}"},{"line_number":3,"context_line":"Q_ML2_TENANT_NETWORK_TYPE\u003d${Q_ML2_TENANT_NETWORK_TYPE:-\"geneve\"}"},{"line_number":4,"context_line":"Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS\u003d${Q_ML2_PLUGIN_GENEVE_TYPE_OPTIONS:-\"vni_ranges\u003d1:65536\"}"},{"line_number":5,"context_line":""},{"line_number":6,"context_line":"# Whether to enable using OVN\u0027s L3 functionality. If this value is disabled,"},{"line_number":7,"context_line":"# OpenStack will use q-l3 functionality."}],"source_content_type":"application/octet-stream","patch_set":4,"id":"3aaa91ec_a56d70ce","line":4,"in_reply_to":"3aaa91ec_e119db7a","updated":"2016-06-28 00:26:29.000000000","message":"cool, even better :)","commit_id":"c6461e61ab5c3de96c2f2421b3794d1689c6194b"}],"doc/source/install.rst":[{"author":{"_account_id":9515,"name":"Matthew Kassawara","email":"mkassawara@gmail.com","username":"ionosphere80"},"change_message_id":"9994501b839f3bad4ec91fe1b08823d5103b8523","unresolved":false,"context_lines":[{"line_number":203,"context_line":"        [ml2_type_geneve]"},{"line_number":204,"context_line":"        ..."},{"line_number":205,"context_line":"        vni_ranges \u003d 1:4095"},{"line_number":206,"context_line":"        max_header_size \u003d 58"},{"line_number":207,"context_line":"        ..."},{"line_number":208,"context_line":""},{"line_number":209,"context_line":"        [securitygroup]"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7aa08908_370bb37e","line":206,"updated":"2016-06-08 22:25:51.000000000","message":"The value of 58 assumes an IPv4 header. For IPv6, add 20. Should this option use a value independent of the IP version and let another component calculate the actual overhead from the tunnel endpoint IP version?","commit_id":"b334c12f2de3b696f8b6957f1323eface81bfc98"},{"author":{"_account_id":1561,"name":"Russell Bryant","email":"rbryant@redhat.com","username":"russellb"},"change_message_id":"3126674d09dc76a2ba2ca69d2a327c99226c6258","unresolved":false,"context_lines":[{"line_number":203,"context_line":"        [ml2_type_geneve]"},{"line_number":204,"context_line":"        ..."},{"line_number":205,"context_line":"        vni_ranges \u003d 1:4095"},{"line_number":206,"context_line":"        max_header_size \u003d 58"},{"line_number":207,"context_line":"        ..."},{"line_number":208,"context_line":""},{"line_number":209,"context_line":"        [securitygroup]"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7aa08908_b4014b07","line":206,"in_reply_to":"7aa08908_370bb37e","updated":"2016-06-09 09:29:34.000000000","message":"This sounds like something we should file a bug for and follow up on.  At least it\u0027s not a regression, but it\u0027s something we should fix.","commit_id":"b334c12f2de3b696f8b6957f1323eface81bfc98"},{"author":{"_account_id":8410,"name":"Richard Theis","email":"rtheis@us.ibm.com","username":"rtheis"},"change_message_id":"430cddd72cd648cc869418df2b22a6895f9b3e17","unresolved":false,"context_lines":[{"line_number":203,"context_line":"        [ml2_type_geneve]"},{"line_number":204,"context_line":"        ..."},{"line_number":205,"context_line":"        vni_ranges \u003d 1:4095"},{"line_number":206,"context_line":"        max_header_size \u003d 58"},{"line_number":207,"context_line":"        ..."},{"line_number":208,"context_line":""},{"line_number":209,"context_line":"        [securitygroup]"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7aa08908_d9c7d662","line":206,"in_reply_to":"7aa08908_b4014b07","updated":"2016-06-09 16:40:50.000000000","message":"It appears that neutron uses this value to calculate the network segment mtu.  However, neutron doesn\u0027t know the IP version at that time.  Should I just add the 20 bytes for a total of 78 to account for the maximum header size?","commit_id":"b334c12f2de3b696f8b6957f1323eface81bfc98"},{"author":{"_account_id":9970,"name":"John Kasperski","email":"jckasper@linux.vnet.ibm.com","username":"jckasper"},"change_message_id":"8b952383125eb66b4ff1fe06ff9e6972283417a6","unresolved":false,"context_lines":[{"line_number":175,"context_line":"        [ml2_type_geneve]"},{"line_number":176,"context_line":"        ..."},{"line_number":177,"context_line":"        vni_ranges \u003d 1:4095"},{"line_number":178,"context_line":"        max_header_size \u003d 58"},{"line_number":179,"context_line":"        ..."},{"line_number":180,"context_line":""},{"line_number":181,"context_line":"        [securitygroup]"}],"source_content_type":"text/x-rst","patch_set":3,"id":"7aa08908_14672e52","line":178,"updated":"2016-06-16 13:33:54.000000000","message":"This defaults to \"50\" here:\n\nhttps://github.com/openstack/neutron/blob/master/neutron/plugins/ml2/drivers/type_geneve.py#L35\n\nShould the default be changed to 58 ?","commit_id":"8162d9d622d851f58db20449a5d9e62d54c2ea4c"},{"author":{"_account_id":8410,"name":"Richard Theis","email":"rtheis@us.ibm.com","username":"rtheis"},"change_message_id":"e99f469d7a2e7807566a6887d0bda3125126b8b1","unresolved":false,"context_lines":[{"line_number":175,"context_line":"        [ml2_type_geneve]"},{"line_number":176,"context_line":"        ..."},{"line_number":177,"context_line":"        vni_ranges \u003d 1:4095"},{"line_number":178,"context_line":"        max_header_size \u003d 58"},{"line_number":179,"context_line":"        ..."},{"line_number":180,"context_line":""},{"line_number":181,"context_line":"        [securitygroup]"}],"source_content_type":"text/x-rst","patch_set":3,"id":"7aa08908_1758d056","line":178,"in_reply_to":"7aa08908_14672e52","updated":"2016-06-16 13:52:10.000000000","message":"I think 50 is the correct default according to http://docs.openstack.org/mitaka/config-reference/networking/networking_options_reference.html.  I believe that OVN adds the extra 8 bytes.","commit_id":"8162d9d622d851f58db20449a5d9e62d54c2ea4c"},{"author":{"_account_id":9970,"name":"John Kasperski","email":"jckasper@linux.vnet.ibm.com","username":"jckasper"},"change_message_id":"37cba131169247b5592edb8c693c5e2fed267b6d","unresolved":false,"context_lines":[{"line_number":175,"context_line":"        [ml2_type_geneve]"},{"line_number":176,"context_line":"        ..."},{"line_number":177,"context_line":"        vni_ranges \u003d 1:4095"},{"line_number":178,"context_line":"        max_header_size \u003d 58"},{"line_number":179,"context_line":"        ..."},{"line_number":180,"context_line":""},{"line_number":181,"context_line":"        [securitygroup]"}],"source_content_type":"text/x-rst","patch_set":3,"id":"7aa08908_1ad75750","line":178,"in_reply_to":"7aa08908_1758d056","updated":"2016-06-16 13:55:37.000000000","message":"8 extra bytes sounds like size of an extra IP option.   Since we don\u0027t know whether IPv6 will be used or not, shouldn\u0027t we assume worst case scenario and always set this to 78 ?","commit_id":"8162d9d622d851f58db20449a5d9e62d54c2ea4c"},{"author":{"_account_id":8410,"name":"Richard Theis","email":"rtheis@us.ibm.com","username":"rtheis"},"change_message_id":"6d985cb04df8aa9c5296698e0a0276dec467bf6a","unresolved":false,"context_lines":[{"line_number":175,"context_line":"        [ml2_type_geneve]"},{"line_number":176,"context_line":"        ..."},{"line_number":177,"context_line":"        vni_ranges \u003d 1:4095"},{"line_number":178,"context_line":"        max_header_size \u003d 58"},{"line_number":179,"context_line":"        ..."},{"line_number":180,"context_line":""},{"line_number":181,"context_line":"        [securitygroup]"}],"source_content_type":"text/x-rst","patch_set":3,"id":"7aa08908_1dc8f174","line":178,"in_reply_to":"7aa08908_1ad75750","updated":"2016-06-16 14:17:04.000000000","message":"https://review.openstack.org/#/c/320121/ is suppose to address this concern so I added a TODO in plugin.sh to revisit this setting once that patch merges.  We should be able to lower the value to 38 and let neutron add 20 or 40 bytes as appropriate.  That said, I\u0027m okay with using 78 bytes now if you\u0027d prefer.","commit_id":"8162d9d622d851f58db20449a5d9e62d54c2ea4c"}],"networking_ovn/ml2/mech_driver.py":[{"author":{"_account_id":11159,"name":"Hong Hui Xiao","email":"honghui_xiao@yeah.net","username":"HongHuiXiao"},"change_message_id":"323361c9a52dd9e11e350a24a4bf8521dfc32d24","unresolved":false,"context_lines":[{"line_number":194,"context_line":"        \"\"\""},{"line_number":195,"context_line":"        network \u003d context.current"},{"line_number":196,"context_line":""},{"line_number":197,"context_line":"        # TODO(rtheis): Add support for multi-provider networks when"},{"line_number":198,"context_line":"        # routed networks are supported."},{"line_number":199,"context_line":"        if self._get_attribute(network, mpnet.SEGMENTS):"},{"line_number":200,"context_line":"            msg \u003d _(\u0027Multi-provider networks are not supported\u0027)"},{"line_number":201,"context_line":"            raise n_exc.InvalidInput(error_message\u003dmsg)"}],"source_content_type":"text/x-python","patch_set":3,"id":"7aa08908_fd5c1e4e","line":198,"range":{"start_line":197,"start_character":0,"end_line":198,"end_character":40},"updated":"2016-06-16 07:02:33.000000000","message":"There is actually muti-provider network[1] before routed network, which means network can be created with multiple segments. During the routed network meeting 1 week ago, we had a discussion about what is routed network. The result is that routed network is not the same as multi-segments network. Routed network is a network with subnets associated with segment. So, for the non routed network(L2 adjacent network), the logic should keep the unchanged, including multi-segments L2 adjacent network.\n\nSo I think this validation here should cover all segments, instead of the first one.\n\n[1] https://blueprints.launchpad.net/python-neutronclient/+spec/neutron-cli-for-creating-multi-segment-network","commit_id":"8162d9d622d851f58db20449a5d9e62d54c2ea4c"},{"author":{"_account_id":11159,"name":"Hong Hui Xiao","email":"honghui_xiao@yeah.net","username":"HongHuiXiao"},"change_message_id":"ebcc379fbd2310a60005d67d050d7f53f81da614","unresolved":false,"context_lines":[{"line_number":194,"context_line":"        \"\"\""},{"line_number":195,"context_line":"        network \u003d context.current"},{"line_number":196,"context_line":""},{"line_number":197,"context_line":"        # TODO(rtheis): Add support for multi-provider networks when"},{"line_number":198,"context_line":"        # routed networks are supported."},{"line_number":199,"context_line":"        if self._get_attribute(network, mpnet.SEGMENTS):"},{"line_number":200,"context_line":"            msg \u003d _(\u0027Multi-provider networks are not supported\u0027)"},{"line_number":201,"context_line":"            raise n_exc.InvalidInput(error_message\u003dmsg)"}],"source_content_type":"text/x-python","patch_set":3,"id":"7aa08908_1baa9d19","line":198,"range":{"start_line":197,"start_character":0,"end_line":198,"end_character":40},"in_reply_to":"7aa08908_b7fd7c0a","updated":"2016-06-17 01:24:49.000000000","message":"Yeah, if the original plugin doesn\u0027t support multi-provider network, I think it is OK to make it be supported with routed network.","commit_id":"8162d9d622d851f58db20449a5d9e62d54c2ea4c"},{"author":{"_account_id":8410,"name":"Richard Theis","email":"rtheis@us.ibm.com","username":"rtheis"},"change_message_id":"233bc5502da1384cb996b8b5d5bc2956efc4a846","unresolved":false,"context_lines":[{"line_number":194,"context_line":"        \"\"\""},{"line_number":195,"context_line":"        network \u003d context.current"},{"line_number":196,"context_line":""},{"line_number":197,"context_line":"        # TODO(rtheis): Add support for multi-provider networks when"},{"line_number":198,"context_line":"        # routed networks are supported."},{"line_number":199,"context_line":"        if self._get_attribute(network, mpnet.SEGMENTS):"},{"line_number":200,"context_line":"            msg \u003d _(\u0027Multi-provider networks are not supported\u0027)"},{"line_number":201,"context_line":"            raise n_exc.InvalidInput(error_message\u003dmsg)"}],"source_content_type":"text/x-python","patch_set":3,"id":"7aa08908_b7fd7c0a","line":198,"range":{"start_line":197,"start_character":0,"end_line":198,"end_character":40},"in_reply_to":"7aa08908_fd5c1e4e","updated":"2016-06-16 13:38:57.000000000","message":"Hi Hong.  That\u0027s correct, but OVN core plugin didn\u0027t support multi-provider networks before and with the switch to ML2, such support wasn\u0027t added.  I figured we could handle multi-provider network support with routed networks support.  Is that okay?","commit_id":"8162d9d622d851f58db20449a5d9e62d54c2ea4c"}]}