)]}'
{"neutron/tests/tempest/api/test_qos.py":[{"author":{"_account_id":6854,"name":"YAMAMOTO Takashi","email":"yamamoto@midokura.com","username":"yamamoto"},"change_message_id":"3593c448727d6b8aec23420963d4fa1c66c15a9b","unresolved":false,"context_lines":[{"line_number":170,"context_line":"        expected_rule_types \u003d ["},{"line_number":171,"context_line":"            qos_consts.RULE_TYPE_BANDWIDTH_LIMIT,"},{"line_number":172,"context_line":"            qos_consts.RULE_TYPE_DSCP_MARKING"},{"line_number":173,"context_line":"        ]"},{"line_number":174,"context_line":"        expected_rule_details \u003d [\u0027type\u0027]"},{"line_number":175,"context_line":""},{"line_number":176,"context_line":"        rule_types \u003d client.list_qos_rule_types()"}],"source_content_type":"text/x-python","patch_set":1,"id":"5f201791_da0cc1a6","line":173,"updated":"2017-06-25 10:44:47.000000000","message":"this seems replacing a wrong assumption with another wrong assumption.\nthe reality is that, the list of expected rule types varies among deployments. we should not hardcode the list here.\nhow about making it an option?","commit_id":"5a83c2b1850fb85b032eb1c43933b85c1f4fd881"},{"author":{"_account_id":11975,"name":"Slawek Kaplonski","email":"skaplons@redhat.com","username":"slaweq"},"change_message_id":"ea4d05b55136272367abc6e649dc99919a46bd4a","unresolved":false,"context_lines":[{"line_number":170,"context_line":"        expected_rule_types \u003d ["},{"line_number":171,"context_line":"            qos_consts.RULE_TYPE_BANDWIDTH_LIMIT,"},{"line_number":172,"context_line":"            qos_consts.RULE_TYPE_DSCP_MARKING"},{"line_number":173,"context_line":"        ]"},{"line_number":174,"context_line":"        expected_rule_details \u003d [\u0027type\u0027]"},{"line_number":175,"context_line":""},{"line_number":176,"context_line":"        rule_types \u003d client.list_qos_rule_types()"}],"source_content_type":"text/x-python","patch_set":1,"id":"5f201791_7dc22b5b","line":173,"in_reply_to":"5f201791_733f13b4","updated":"2017-06-26 16:45:36.000000000","message":"YAMAMOTO: so maybe change it to test if keys returned from server are proper and not test values which are dependent of deployment. What You think about such solution?","commit_id":"5a83c2b1850fb85b032eb1c43933b85c1f4fd881"},{"author":{"_account_id":6854,"name":"YAMAMOTO Takashi","email":"yamamoto@midokura.com","username":"yamamoto"},"change_message_id":"953c3c8bab233dbdf07305c8923ad6353c375885","unresolved":false,"context_lines":[{"line_number":170,"context_line":"        expected_rule_types \u003d ["},{"line_number":171,"context_line":"            qos_consts.RULE_TYPE_BANDWIDTH_LIMIT,"},{"line_number":172,"context_line":"            qos_consts.RULE_TYPE_DSCP_MARKING"},{"line_number":173,"context_line":"        ]"},{"line_number":174,"context_line":"        expected_rule_details \u003d [\u0027type\u0027]"},{"line_number":175,"context_line":""},{"line_number":176,"context_line":"        rule_types \u003d client.list_qos_rule_types()"}],"source_content_type":"text/x-python","patch_set":1,"id":"5f201791_9a1089aa","line":173,"in_reply_to":"5f201791_7afb6d78","updated":"2017-06-25 11:07:57.000000000","message":"then i guess the \"different patch\" should go in first, to avoid hurting deployments other than neutron gate.","commit_id":"5a83c2b1850fb85b032eb1c43933b85c1f4fd881"},{"author":{"_account_id":11975,"name":"Slawek Kaplonski","email":"skaplons@redhat.com","username":"slaweq"},"change_message_id":"5ce29dc1446bb4c28e331da4c6ae851452db8975","unresolved":false,"context_lines":[{"line_number":170,"context_line":"        expected_rule_types \u003d ["},{"line_number":171,"context_line":"            qos_consts.RULE_TYPE_BANDWIDTH_LIMIT,"},{"line_number":172,"context_line":"            qos_consts.RULE_TYPE_DSCP_MARKING"},{"line_number":173,"context_line":"        ]"},{"line_number":174,"context_line":"        expected_rule_details \u003d [\u0027type\u0027]"},{"line_number":175,"context_line":""},{"line_number":176,"context_line":"        rule_types \u003d client.list_qos_rule_types()"}],"source_content_type":"text/x-python","patch_set":1,"id":"5f201791_eb954c41","line":173,"in_reply_to":"5f201791_7dc22b5b","updated":"2017-06-26 21:43:40.000000000","message":"As I discussed with Kevin Benton on IRC meeting today: http://eavesdrop.openstack.org/meetings/networking/2017/networking.2017-06-26-21.17.log.html#l-53\nI will update this test to check only returned keys and not test values which depends on deployment","commit_id":"5a83c2b1850fb85b032eb1c43933b85c1f4fd881"},{"author":{"_account_id":11975,"name":"Slawek Kaplonski","email":"skaplons@redhat.com","username":"slaweq"},"change_message_id":"895a51291c5582b3fc9913669e42cff906856c99","unresolved":false,"context_lines":[{"line_number":170,"context_line":"        expected_rule_types \u003d ["},{"line_number":171,"context_line":"            qos_consts.RULE_TYPE_BANDWIDTH_LIMIT,"},{"line_number":172,"context_line":"            qos_consts.RULE_TYPE_DSCP_MARKING"},{"line_number":173,"context_line":"        ]"},{"line_number":174,"context_line":"        expected_rule_details \u003d [\u0027type\u0027]"},{"line_number":175,"context_line":""},{"line_number":176,"context_line":"        rule_types \u003d client.list_qos_rule_types()"}],"source_content_type":"text/x-python","patch_set":1,"id":"5f201791_ee345e95","line":173,"in_reply_to":"5f201791_9a1089aa","updated":"2017-06-25 13:55:03.000000000","message":"did You think about something similar to \"available_type_drivers\" list in tempest config?\nWe can add something similar with available_mechanism_drivers but I\u0027m not sure if this will solve problem for other deployments (not related to ML2)","commit_id":"5a83c2b1850fb85b032eb1c43933b85c1f4fd881"},{"author":{"_account_id":6854,"name":"YAMAMOTO Takashi","email":"yamamoto@midokura.com","username":"yamamoto"},"change_message_id":"e5cbd0f0efc296df9136f98b5fbbe9aa2a8d9952","unresolved":false,"context_lines":[{"line_number":170,"context_line":"        expected_rule_types \u003d ["},{"line_number":171,"context_line":"            qos_consts.RULE_TYPE_BANDWIDTH_LIMIT,"},{"line_number":172,"context_line":"            qos_consts.RULE_TYPE_DSCP_MARKING"},{"line_number":173,"context_line":"        ]"},{"line_number":174,"context_line":"        expected_rule_details \u003d [\u0027type\u0027]"},{"line_number":175,"context_line":""},{"line_number":176,"context_line":"        rule_types \u003d client.list_qos_rule_types()"}],"source_content_type":"text/x-python","patch_set":1,"id":"5f201791_733f13b4","line":173,"in_reply_to":"5f201791_ae10a6a9","updated":"2017-06-26 05:09:17.000000000","message":"i suspect we should not verify that details of deployments in tempest tests.\nwe have a similar issue for agent extensions.\nbug 1509590\nbug 1699199\ni guess we need to find compromises in case-by-case basis.","commit_id":"5a83c2b1850fb85b032eb1c43933b85c1f4fd881"},{"author":{"_account_id":11975,"name":"Slawek Kaplonski","email":"skaplons@redhat.com","username":"slaweq"},"change_message_id":"7b77f2d95e83e994eadc7cb4a9c50e467a409d4a","unresolved":false,"context_lines":[{"line_number":170,"context_line":"        expected_rule_types \u003d ["},{"line_number":171,"context_line":"            qos_consts.RULE_TYPE_BANDWIDTH_LIMIT,"},{"line_number":172,"context_line":"            qos_consts.RULE_TYPE_DSCP_MARKING"},{"line_number":173,"context_line":"        ]"},{"line_number":174,"context_line":"        expected_rule_details \u003d [\u0027type\u0027]"},{"line_number":175,"context_line":""},{"line_number":176,"context_line":"        rule_types \u003d client.list_qos_rule_types()"}],"source_content_type":"text/x-python","patch_set":1,"id":"5f201791_ae10a6a9","line":173,"in_reply_to":"5f201791_cea8828d","updated":"2017-06-25 14:50:38.000000000","message":"yes, but this will not fix \"future\" issue. For example similar case will be in my new patch https://review.openstack.org/#/c/475260/5/neutron/tests/tempest/api/test_qos.py\nWhat You think about that?","commit_id":"5a83c2b1850fb85b032eb1c43933b85c1f4fd881"},{"author":{"_account_id":11975,"name":"Slawek Kaplonski","email":"skaplons@redhat.com","username":"slaweq"},"change_message_id":"7f5c71d04321537f2e34e2ff9d78eeabc2c6c956","unresolved":false,"context_lines":[{"line_number":170,"context_line":"        expected_rule_types \u003d ["},{"line_number":171,"context_line":"            qos_consts.RULE_TYPE_BANDWIDTH_LIMIT,"},{"line_number":172,"context_line":"            qos_consts.RULE_TYPE_DSCP_MARKING"},{"line_number":173,"context_line":"        ]"},{"line_number":174,"context_line":"        expected_rule_details \u003d [\u0027type\u0027]"},{"line_number":175,"context_line":""},{"line_number":176,"context_line":"        rule_types \u003d client.list_qos_rule_types()"}],"source_content_type":"text/x-python","patch_set":1,"id":"5f201791_7afb6d78","line":173,"in_reply_to":"5f201791_da0cc1a6","updated":"2017-06-25 10:48:56.000000000","message":"in fact this patch is about fixing test which is broken currently and this fix is done in lines below (changed again in PS2)\n\nThis expected rule types list is changed just to make fixed test passing. It not changes anything in test logic.\n\nI agree that we should make it more deployment independent but IMHO it should be different patch for that.","commit_id":"5a83c2b1850fb85b032eb1c43933b85c1f4fd881"},{"author":{"_account_id":6854,"name":"YAMAMOTO Takashi","email":"yamamoto@midokura.com","username":"yamamoto"},"change_message_id":"d0e45b3dcc46d8bf593a5374091425ba3aa8ff4e","unresolved":false,"context_lines":[{"line_number":170,"context_line":"        expected_rule_types \u003d ["},{"line_number":171,"context_line":"            qos_consts.RULE_TYPE_BANDWIDTH_LIMIT,"},{"line_number":172,"context_line":"            qos_consts.RULE_TYPE_DSCP_MARKING"},{"line_number":173,"context_line":"        ]"},{"line_number":174,"context_line":"        expected_rule_details \u003d [\u0027type\u0027]"},{"line_number":175,"context_line":""},{"line_number":176,"context_line":"        rule_types \u003d client.list_qos_rule_types()"}],"source_content_type":"text/x-python","patch_set":1,"id":"5f201791_cea8828d","line":173,"in_reply_to":"5f201791_ee345e95","updated":"2017-06-25 14:19:56.000000000","message":"i was thinking about something more direct.\neg. available_qos_rule_types.","commit_id":"5a83c2b1850fb85b032eb1c43933b85c1f4fd881"},{"author":{"_account_id":1653,"name":"garyk","email":"gkotton@vmware.com","username":"garyk"},"change_message_id":"dbb9f54c91d709efb2be649c3abfbcea88651afc","unresolved":false,"context_lines":[{"line_number":158,"context_line":"        # List supported rule types"},{"line_number":159,"context_line":"        # TODO(QoS): since in gate we run both ovs and linuxbridge ml2 drivers,"},{"line_number":160,"context_line":"        # and since both backend supports same QoS rule types, the list of"},{"line_number":161,"context_line":"        # expected rule types should contains rules supported by both backends:"},{"line_number":162,"context_line":"        # dscp marking rule and bandwidth limit rule"},{"line_number":163,"context_line":"        # Once one of those agents receive support for another QoS rule type,"},{"line_number":164,"context_line":"        #  the list of expected rule types will change."}],"source_content_type":"text/x-python","patch_set":2,"id":"5f201791_9a488987","line":161,"range":{"start_line":161,"start_character":37,"end_line":161,"end_character":45},"updated":"2017-06-25 11:57:52.000000000","message":"contain","commit_id":"c9bf3554a88c1238e8daf936035a07e06e42a764"}]}