)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":24245,"name":"Harald Jensås","email":"hjensas@redhat.com","username":"harald.jensas"},"change_message_id":"ad8f34dd076dcfbc762d98749e6eccd96c70fd79","unresolved":false,"context_lines":[{"line_number":9,"context_line":"Ironic added support for physical network awareness [1] in the pike"},{"line_number":10,"context_line":"release. With multiple physical networks it is possible that each"},{"line_number":11,"context_line":"physical network extends over a subset of the devices in a network."},{"line_number":12,"context_line":"Further, the segmentation ID space on different physical networks may"},{"line_number":13,"context_line":"overlap."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"This change adds support for configuration of a set of physical networks"},{"line_number":16,"context_line":"on each switch. On neutron network creation and deletion, NGS will only"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":3,"id":"9f91af0f_1e702fa5","line":13,"range":{"start_line":12,"start_character":0,"end_line":13,"end_character":8},"updated":"2018-01-04 12:30:36.000000000","message":"Question: \n Do we somehow handle this? Can we? Segmentation ID is essentially VLAN id\u0027s right? If we add physnet1 and physnet2 on the same device, and both physnet uses vlan id 100 we are effectively bridging the networks togheter?","commit_id":"e795a7c844584ba9147bbc0ba6af4a3a321876df"},{"author":{"_account_id":14826,"name":"Mark Goddard","email":"markgoddard86@gmail.com","username":"mgoddard"},"change_message_id":"fffff6e9c3ea8bcc597541447801eaee715a0f4b","unresolved":false,"context_lines":[{"line_number":9,"context_line":"Ironic added support for physical network awareness [1] in the pike"},{"line_number":10,"context_line":"release. With multiple physical networks it is possible that each"},{"line_number":11,"context_line":"physical network extends over a subset of the devices in a network."},{"line_number":12,"context_line":"Further, the segmentation ID space on different physical networks may"},{"line_number":13,"context_line":"overlap."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"This change adds support for configuration of a set of physical networks"},{"line_number":16,"context_line":"on each switch. On neutron network creation and deletion, NGS will only"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":3,"id":"9f91af0f_5ed2375e","line":13,"range":{"start_line":12,"start_character":0,"end_line":13,"end_character":8},"in_reply_to":"9f91af0f_1e702fa5","updated":"2018-01-04 12:38:55.000000000","message":"I don\u0027t think we can avoid this - if you use the same VLAN ID on two physnets on the same switch, they will be bridged. My advice would be to not do it :)","commit_id":"e795a7c844584ba9147bbc0ba6af4a3a321876df"},{"author":{"_account_id":24245,"name":"Harald Jensås","email":"hjensas@redhat.com","username":"harald.jensas"},"change_message_id":"a567b49097a8110c5803545991cbb289e4e4c34c","unresolved":false,"context_lines":[{"line_number":9,"context_line":"Ironic added support for physical network awareness [1] in the pike"},{"line_number":10,"context_line":"release. With multiple physical networks it is possible that each"},{"line_number":11,"context_line":"physical network extends over a subset of the devices in a network."},{"line_number":12,"context_line":"Further, the segmentation ID space on different physical networks may"},{"line_number":13,"context_line":"overlap."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"This change adds support for configuration of a set of physical networks"},{"line_number":16,"context_line":"on each switch. On neutron network creation and deletion, NGS will only"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":3,"id":"9f91af0f_b298571b","line":13,"range":{"start_line":12,"start_character":0,"end_line":13,"end_character":8},"in_reply_to":"9f91af0f_5ed2375e","updated":"2018-01-04 14:13:19.000000000","message":"Would it make sense to implement a safety in create_network_precommit in a follow on patch? Check for an existing VLAN with the segmentation ID on the switch, and raise an exception in case it is already there? Ideally we would be able to store the physical_network a vlan was created for ...","commit_id":"e795a7c844584ba9147bbc0ba6af4a3a321876df"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"d81a4ad1cb99dd4af1a295ded1289cf7d4afeb97","unresolved":false,"context_lines":[{"line_number":9,"context_line":"Ironic added support for physical network awareness [1] in the pike"},{"line_number":10,"context_line":"release. With multiple physical networks it is possible that each"},{"line_number":11,"context_line":"physical network extends over a subset of the devices in a network."},{"line_number":12,"context_line":"Further, the segmentation ID space on different physical networks may"},{"line_number":13,"context_line":"overlap."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"This change adds support for configuration of a set of physical networks"},{"line_number":16,"context_line":"on each switch. On neutron network creation and deletion, NGS will only"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":3,"id":"9f91af0f_ab0e8b2f","line":13,"range":{"start_line":12,"start_character":0,"end_line":13,"end_character":8},"in_reply_to":"9f91af0f_a8ad3132","updated":"2018-01-05 14:31:54.000000000","message":"So thinking about it even further, in the physical world it would/could be be a valid configuration depending on things like spanning tree configuration inside the vlans. I think a good route could be a sanity checking tool that looks for overlaps, since neutron would have to be queried. Just a thought at least. A \"n-g-s monitoring\" script or something. Just an idea.","commit_id":"e795a7c844584ba9147bbc0ba6af4a3a321876df"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"efc7e0694c07fb7f30e9077dd386c7c1ee78eec1","unresolved":false,"context_lines":[{"line_number":9,"context_line":"Ironic added support for physical network awareness [1] in the pike"},{"line_number":10,"context_line":"release. With multiple physical networks it is possible that each"},{"line_number":11,"context_line":"physical network extends over a subset of the devices in a network."},{"line_number":12,"context_line":"Further, the segmentation ID space on different physical networks may"},{"line_number":13,"context_line":"overlap."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"This change adds support for configuration of a set of physical networks"},{"line_number":16,"context_line":"on each switch. On neutron network creation and deletion, NGS will only"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":3,"id":"9f91af0f_a8ad3132","line":13,"range":{"start_line":12,"start_character":0,"end_line":13,"end_character":8},"in_reply_to":"9f91af0f_b298571b","updated":"2018-01-05 14:29:42.000000000","message":"But the differing physnets, at least as I understand it, should be sufficiently delineated by the operator. I think in reality, it is the same physnet to the same switch then it is a invalid configuration in the basic setup of the environment. I\u0027m not sure we should put guarding logic to prevent that, because it could also be an intent.","commit_id":"e795a7c844584ba9147bbc0ba6af4a3a321876df"}],"networking_generic_switch/generic_switch_mech.py":[{"author":{"_account_id":14525,"name":"Vasyl Saienko","email":"vsaienko@mirantis.com","username":"vsaienko"},"change_message_id":"9d33b8350d024e3343be813a961469e3df4a0083","unresolved":false,"context_lines":[{"line_number":504,"context_line":"            return"},{"line_number":505,"context_line":"        switch_info \u003d local_link_information[0].get(\u0027switch_info\u0027)"},{"line_number":506,"context_line":"        switch_id \u003d local_link_information[0].get(\u0027switch_id\u0027)"},{"line_number":507,"context_line":"        switch \u003d device_utils.get_switch_device("},{"line_number":508,"context_line":"            self.switches, switch_info\u003dswitch_info,"},{"line_number":509,"context_line":"            ngs_mac_address\u003dswitch_id)"},{"line_number":510,"context_line":"        if not switch:"}],"source_content_type":"text/x-python","patch_set":2,"id":"df87a7cf_34bc25ba","line":507,"updated":"2017-12-06 10:25:14.000000000","message":"wondered if we should add the same restriction here, but if we already bound port it means physnet was in the list and it may blow here only when configuration of physnets was changed after port was bound.","commit_id":"4592731cefd844c591d559445cd0977864c23f1d"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"fc5f137b2bc8a62ef043def9e3b4c7cbdee49bf7","unresolved":false,"context_lines":[{"line_number":437,"context_line":"                          \"not on physical network %(physnet)\","},{"line_number":438,"context_line":"                          {\u0027port_id\u0027: port[\u0027id\u0027], \u0027device\u0027: switch_info,"},{"line_number":439,"context_line":"                           \u0027physnet\u0027: physnet})"},{"line_number":440,"context_line":"                return"},{"line_number":441,"context_line":"            port_id \u003d local_link_information[0].get(\u0027port_id\u0027)"},{"line_number":442,"context_line":"            segments \u003d context.segments_to_bind"},{"line_number":443,"context_line":"            segmentation_id \u003d segments[0].get(\u0027segmentation_id\u0027)"}],"source_content_type":"text/x-python","patch_set":3,"id":"9f91af0f_6550b4d5","line":440,"updated":"2018-01-05 12:56:35.000000000","message":"is it fine to not raise anything here? how will users know about the error?","commit_id":"e795a7c844584ba9147bbc0ba6af4a3a321876df"},{"author":{"_account_id":14826,"name":"Mark Goddard","email":"markgoddard86@gmail.com","username":"mgoddard"},"change_message_id":"781d2784c157194a2f5a05ae4b216645f118302a","unresolved":false,"context_lines":[{"line_number":437,"context_line":"                          \"not on physical network %(physnet)\","},{"line_number":438,"context_line":"                          {\u0027port_id\u0027: port[\u0027id\u0027], \u0027device\u0027: switch_info,"},{"line_number":439,"context_line":"                           \u0027physnet\u0027: physnet})"},{"line_number":440,"context_line":"                return"},{"line_number":441,"context_line":"            port_id \u003d local_link_information[0].get(\u0027port_id\u0027)"},{"line_number":442,"context_line":"            segments \u003d context.segments_to_bind"},{"line_number":443,"context_line":"            segmentation_id \u003d segments[0].get(\u0027segmentation_id\u0027)"}],"source_content_type":"text/x-python","patch_set":3,"id":"9f91af0f_3a92bf9f","line":440,"in_reply_to":"9f91af0f_2835a1f7","updated":"2018-01-05 18:01:45.000000000","message":"It\u0027s a good question. I think Julia\u0027s correct though - neutron logs and ignores exceptions from mech drivers in bind_port, so we may as well log the error on our terms.","commit_id":"e795a7c844584ba9147bbc0ba6af4a3a321876df"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"efc7e0694c07fb7f30e9077dd386c7c1ee78eec1","unresolved":false,"context_lines":[{"line_number":437,"context_line":"                          \"not on physical network %(physnet)\","},{"line_number":438,"context_line":"                          {\u0027port_id\u0027: port[\u0027id\u0027], \u0027device\u0027: switch_info,"},{"line_number":439,"context_line":"                           \u0027physnet\u0027: physnet})"},{"line_number":440,"context_line":"                return"},{"line_number":441,"context_line":"            port_id \u003d local_link_information[0].get(\u0027port_id\u0027)"},{"line_number":442,"context_line":"            segments \u003d context.segments_to_bind"},{"line_number":443,"context_line":"            segmentation_id \u003d segments[0].get(\u0027segmentation_id\u0027)"}],"source_content_type":"text/x-python","patch_set":3,"id":"9f91af0f_2835a1f7","line":440,"in_reply_to":"9f91af0f_6550b4d5","updated":"2018-01-05 14:29:42.000000000","message":"I don\u0027t think we really can because I\u0027m not sure raising an error here gets back to neutron to fail the port being brought up. Besides, depending on the port, the user might have no real insight into that there has been a failure until they were to look at neutron. :\\","commit_id":"e795a7c844584ba9147bbc0ba6af4a3a321876df"}]}