)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":6788,"name":"Rossella Sblendido","email":"rsblendido@suse.com","username":"rossella-o"},"change_message_id":"ed2b601d3b794f4c6078b3ecedf82642236a3d7a","unresolved":false,"context_lines":[{"line_number":6,"context_line":""},{"line_number":7,"context_line":"ovs: set device MTU after it\u0027s moved into a namespace"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"For some unclear reason, Open vSwitch forbids to raise MTU for devices"},{"line_number":10,"context_line":"that are in the same namespace with the bridge, but allows that same"},{"line_number":11,"context_line":"operation once the device in question is moved into another namespace."},{"line_number":12,"context_line":""}],"source_content_type":"text/x-gerrit-commit-message","patch_set":4,"id":"7aa08908_0e840e5b","line":9,"range":{"start_line":9,"start_character":0,"end_line":9,"end_character":23},"updated":"2016-06-17 08:02:54.000000000","message":"from what I see from the mail thread, the reason is clear now, can you add that info in the commit message please? I think we should state here that we are exploiting an OVS bug.","commit_id":"768bbc6da5f22fb67e8880a9423a64529efb990b"},{"author":{"_account_id":13252,"name":"Dr. Jens Harbott","display_name":"Jens Harbott (frickler)","email":"frickler@offenerstapel.de","username":"jrosenboom"},"change_message_id":"35546384e9a4421a7b02d6348f2adfa7c077e944","unresolved":false,"context_lines":[{"line_number":17,"context_line":""},{"line_number":18,"context_line":"This effectively lowers MTU for router interfaces from what is supported"},{"line_number":19,"context_line":"by underlying physical interfaces. With that, frames from an external"},{"line_number":20,"context_line":"network of infrastructure MTU size will not be served by affected"},{"line_number":21,"context_line":"routers."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"This bug affects all systems where router or dhcp ports are served on"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":4,"id":"7aa08908_15b0e48a","line":20,"updated":"2016-06-10 05:27:40.000000000","message":"According to my tests, this is not quite correct, a lower MTU will not imply a lower MRU. I.e. the router will still receive packets with a size larger than the MTU on that port, but outgoing packets will be fragmented.","commit_id":"768bbc6da5f22fb67e8880a9423a64529efb990b"},{"author":{"_account_id":9656,"name":"Ihar Hrachyshka","email":"ihrachys@redhat.com","username":"ihrachys","status":"Red Hat Networking Systems Engineer"},"change_message_id":"796b56065c30157b0234943698c871bd3141c6e3","unresolved":false,"context_lines":[{"line_number":17,"context_line":""},{"line_number":18,"context_line":"This effectively lowers MTU for router interfaces from what is supported"},{"line_number":19,"context_line":"by underlying physical interfaces. With that, frames from an external"},{"line_number":20,"context_line":"network of infrastructure MTU size will not be served by affected"},{"line_number":21,"context_line":"routers."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"This bug affects all systems where router or dhcp ports are served on"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":4,"id":"7aa08908_0872a2b0","line":20,"in_reply_to":"7aa08908_15b0e48a","updated":"2016-06-15 16:16:34.000000000","message":"It\u0027s only valid for one path - from external network into instances. But if your instances will produce MTU sized frames, they will not reach the router, because the qr leg that lays on the data path for those frames has wrong (lower) mtu value.","commit_id":"768bbc6da5f22fb67e8880a9423a64529efb990b"}],"neutron/agent/linux/interface.py":[{"author":{"_account_id":9656,"name":"Ihar Hrachyshka","email":"ihrachys@redhat.com","username":"ihrachys","status":"Red Hat Networking Systems Engineer"},"change_message_id":"b75dee3a757a7d772651cc16724c38696751e005","unresolved":false,"context_lines":[{"line_number":343,"context_line":""},{"line_number":344,"context_line":"        ns_dev.link.set_address(mac_address)"},{"line_number":345,"context_line":""},{"line_number":346,"context_line":"        # Add an interface created by ovs to the namespace."},{"line_number":347,"context_line":"        if not self.conf.ovs_use_veth and namespace:"},{"line_number":348,"context_line":"            namespace_obj \u003d ip.ensure_namespace(namespace)"},{"line_number":349,"context_line":"            namespace_obj.add_device_to_namespace(ns_dev)"}],"source_content_type":"text/x-python","patch_set":1,"id":"7aa08908_690b7e2c","line":346,"updated":"2016-06-09 13:29:29.000000000","message":"Need to add a comment here describing the specific order we pick to set MTU.","commit_id":"199f7cd069255ee0eeab77be23bdd92c19ebcbb4"},{"author":{"_account_id":9656,"name":"Ihar Hrachyshka","email":"ihrachys@redhat.com","username":"ihrachys","status":"Red Hat Networking Systems Engineer"},"change_message_id":"5d7ec08b3820fb4244022ce360b4d8b814adddf9","unresolved":false,"context_lines":[{"line_number":351,"context_line":"        mtu \u003d self.conf.network_device_mtu or mtu"},{"line_number":352,"context_line":"        if mtu:"},{"line_number":353,"context_line":"            ns_dev.link.set_mtu(mtu)"},{"line_number":354,"context_line":"            if self.conf.ovs_use_veth:"},{"line_number":355,"context_line":"                root_dev.link.set_mtu(mtu)"},{"line_number":356,"context_line":"        else:"},{"line_number":357,"context_line":"            LOG.warning(_LW(\"No MTU configured for port %s\"), port_id)"}],"source_content_type":"text/x-python","patch_set":1,"id":"7aa08908_b1f14498","line":354,"updated":"2016-06-09 15:57:06.000000000","message":"Other drivers may require a similar order reshuffle too.","commit_id":"199f7cd069255ee0eeab77be23bdd92c19ebcbb4"}]}
