)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":1131,"name":"Brian Haley","email":"haleyb.dev@gmail.com","username":"brian-haley"},"change_message_id":"ebeb2eeea04ea7e4ba86cb9c61936ac2fd6d30fb","unresolved":false,"context_lines":[{"line_number":19,"context_line":"security issue."},{"line_number":20,"context_line":""},{"line_number":21,"context_line":"Change-Id: Ibf858c7fdf7a822a30e4a0c4722d70fd272741b6"},{"line_number":22,"context_line":"Related-bug: #1630832"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"5f93b717_dea3d84e","line":22,"updated":"2018-01-25 19:42:31.000000000","message":"I\u0027m going to open a new bug as well instead of use the one for the netlink contrack RFE.","commit_id":"0b75e33eb4cc8d0a0d3c97579c45e676c7f04204"},{"author":{"_account_id":11975,"name":"Slawek Kaplonski","email":"skaplons@redhat.com","username":"slaweq"},"change_message_id":"12a7bb1115c345b7b6545299aede38815cdceddd","unresolved":false,"context_lines":[{"line_number":18,"context_line":"NAT rules will have been removed, it should not pose a"},{"line_number":19,"context_line":"security issue."},{"line_number":20,"context_line":""},{"line_number":21,"context_line":"A new config option in the [security] section of ml2_conf.ini,"},{"line_number":22,"context_line":"`conntrack_workers`, can be set as desired, or set to zero to"},{"line_number":23,"context_line":"restore the previous behavior."},{"line_number":24,"context_line":""}],"source_content_type":"text/x-gerrit-commit-message","patch_set":3,"id":"3fa0c359_c8349ce1","line":21,"range":{"start_line":21,"start_character":27,"end_line":21,"end_character":37},"updated":"2018-02-05 11:15:09.000000000","message":"securitygroup section","commit_id":"3294e923e7449d207b973de78716965a1a4770de"},{"author":{"_account_id":1131,"name":"Brian Haley","email":"haleyb.dev@gmail.com","username":"brian-haley"},"change_message_id":"8c09195e72cf1dd4f7c02cd11620df4198de75a4","unresolved":false,"context_lines":[{"line_number":18,"context_line":"NAT rules will have been removed, it should not pose a"},{"line_number":19,"context_line":"security issue."},{"line_number":20,"context_line":""},{"line_number":21,"context_line":"A new config option in the [security] section of ml2_conf.ini,"},{"line_number":22,"context_line":"`conntrack_workers`, can be set as desired, or set to zero to"},{"line_number":23,"context_line":"restore the previous behavior."},{"line_number":24,"context_line":""}],"source_content_type":"text/x-gerrit-commit-message","patch_set":3,"id":"3fa0c359_e41e556f","line":21,"range":{"start_line":21,"start_character":27,"end_line":21,"end_character":37},"in_reply_to":"3fa0c359_c8349ce1","updated":"2018-02-05 20:36:16.000000000","message":"Done","commit_id":"3294e923e7449d207b973de78716965a1a4770de"},{"author":{"_account_id":4694,"name":"Miguel Lavalle","email":"miguel@mlavalle.com","username":"minsel"},"change_message_id":"247d8b2c4028ac947208e871c3d23e8d7f43c066","unresolved":false,"context_lines":[{"line_number":9,"context_line":"With a large number of instances and/or security group rules,"},{"line_number":10,"context_line":"conntrack updates when ports are removed or rules are changed"},{"line_number":11,"context_line":"can take a long time to process.  By enqueuing these to a set"},{"line_number":12,"context_line":"or worker threads, the agent can continue with other work while"},{"line_number":13,"context_line":"they are processed in the background."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"This is a change in behavior in the agent since it could"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":4,"id":"3fa0c359_015a3af2","line":12,"range":{"start_line":12,"start_character":0,"end_line":12,"end_character":2},"updated":"2018-02-16 21:04:21.000000000","message":"of","commit_id":"58b8a437c09bbd83fcb610b826ed8cabcf56a384"},{"author":{"_account_id":11975,"name":"Slawek Kaplonski","email":"skaplons@redhat.com","username":"slaweq"},"change_message_id":"4b97c99ada3aeb9b44c802a44d6c7858d51b9d63","unresolved":false,"context_lines":[{"line_number":24,"context_line":"is (1 + \u003cnumber of cpus\u003e) // 2."},{"line_number":25,"context_line":""},{"line_number":26,"context_line":"Change-Id: Ibf858c7fdf7a822a30e4a0c4722d70fd272741b6"},{"line_number":27,"context_line":"Partial-bug: #1745468"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":4,"id":"3fa0c359_2c5414ca","line":27,"updated":"2018-02-06 10:17:42.000000000","message":"Just one more question (out of curiosity): what else can/should be done to close this bug report?","commit_id":"58b8a437c09bbd83fcb610b826ed8cabcf56a384"},{"author":{"_account_id":11975,"name":"Slawek Kaplonski","email":"skaplons@redhat.com","username":"slaweq"},"change_message_id":"eee045fa081362004d58db8812ebd4e5cf10d379","unresolved":false,"context_lines":[{"line_number":24,"context_line":"is (1 + \u003cnumber of cpus\u003e) // 2."},{"line_number":25,"context_line":""},{"line_number":26,"context_line":"Change-Id: Ibf858c7fdf7a822a30e4a0c4722d70fd272741b6"},{"line_number":27,"context_line":"Partial-bug: #1745468"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":4,"id":"3fa0c359_cc2c9049","line":27,"in_reply_to":"3fa0c359_2c5414ca","updated":"2018-02-06 10:21:51.000000000","message":"ok, forget my question. I already found it :) Sorry","commit_id":"58b8a437c09bbd83fcb610b826ed8cabcf56a384"},{"author":{"_account_id":1131,"name":"Brian Haley","email":"haleyb.dev@gmail.com","username":"brian-haley"},"change_message_id":"e865f05529e5e654c52017b5cb455de1c4e3a111","unresolved":false,"context_lines":[{"line_number":9,"context_line":"With a large number of instances and/or security group rules,"},{"line_number":10,"context_line":"conntrack updates when ports are removed or rules are changed"},{"line_number":11,"context_line":"can take a long time to process.  By enqueuing these to a set"},{"line_number":12,"context_line":"or worker threads, the agent can continue with other work while"},{"line_number":13,"context_line":"they are processed in the background."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"This is a change in behavior in the agent since it could"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":6,"id":"3fa0c359_7db873e4","line":12,"range":{"start_line":12,"start_character":0,"end_line":12,"end_character":2},"updated":"2018-02-16 21:08:27.000000000","message":"if i have to respin i will change this to \"of\" :)","commit_id":"65a81623fc0377b26d2d5800607f7c3acc08c45a"}],"neutron/agent/linux/ip_conntrack.py":[{"author":{"_account_id":1131,"name":"Brian Haley","email":"haleyb.dev@gmail.com","username":"brian-haley"},"change_message_id":"e98a378917634298fdb10fddbe80bfc83436223f","unresolved":false,"context_lines":[{"line_number":89,"context_line":""},{"line_number":90,"context_line":"    def _process_queue_loop(self):"},{"line_number":91,"context_line":"        LOG.debug(\"Starting ipconntrack _process_queue_loop()\")"},{"line_number":92,"context_line":"        pool \u003d eventlet.GreenPool(size\u003d8)"},{"line_number":93,"context_line":"        while True:"},{"line_number":94,"context_line":"            pool.spawn_n(self._process_queue)"},{"line_number":95,"context_line":""}],"source_content_type":"text/x-python","patch_set":1,"id":"5f93b717_ca40fb0a","line":92,"updated":"2018-01-25 14:00:50.000000000","message":"I am tempted to make the number of workers here configurable, with zero meaning just process the update immediately in the calling thread as we currently do.","commit_id":"0b75e33eb4cc8d0a0d3c97579c45e676c7f04204"},{"author":{"_account_id":1131,"name":"Brian Haley","email":"haleyb.dev@gmail.com","username":"brian-haley"},"change_message_id":"accefcfea1dca4200c967b184a12ba49b710d298","unresolved":false,"context_lines":[{"line_number":102,"context_line":"        for update in self._queue.updates():"},{"line_number":103,"context_line":"            LOG.debug(\"Starting ipconntrack _process_queue() update\")"},{"line_number":104,"context_line":"            self._process_update(update)"},{"line_number":105,"context_line":"            LOG.debug(\"Finished ipconntrack _process_queue() update\")"},{"line_number":106,"context_line":""},{"line_number":107,"context_line":"    def _process_update(self, update):"},{"line_number":108,"context_line":"        if update.remote_ips:"}],"source_content_type":"text/x-python","patch_set":2,"id":"5f93b717_3cefbae9","line":105,"updated":"2018-02-01 17:10:14.000000000","message":"I\u0027ve found while benchmarking that these log messages aren\u0027t very useful, so will remove them in the next PS.","commit_id":"9bc7106738f56735be38159b5a4f3bb02b43a687"},{"author":{"_account_id":11975,"name":"Slawek Kaplonski","email":"skaplons@redhat.com","username":"slaweq"},"change_message_id":"12a7bb1115c345b7b6545299aede38815cdceddd","unresolved":false,"context_lines":[{"line_number":94,"context_line":""},{"line_number":95,"context_line":"    def _process_queue_loop(self):"},{"line_number":96,"context_line":"        LOG.debug(\"Starting ipconntrack _process_queue_loop()\")"},{"line_number":97,"context_line":"        pool \u003d eventlet.GreenPool(size\u003d8)"},{"line_number":98,"context_line":"        while True:"},{"line_number":99,"context_line":"            pool.spawn_n(self._process_queue)"},{"line_number":100,"context_line":""}],"source_content_type":"text/x-python","patch_set":3,"id":"3fa0c359_9cbc9abc","line":97,"range":{"start_line":97,"start_character":34,"end_line":97,"end_character":40},"updated":"2018-02-05 11:15:09.000000000","message":"why hardcoded value and why \"8\"?","commit_id":"3294e923e7449d207b973de78716965a1a4770de"},{"author":{"_account_id":1131,"name":"Brian Haley","email":"haleyb.dev@gmail.com","username":"brian-haley"},"change_message_id":"8c09195e72cf1dd4f7c02cd11620df4198de75a4","unresolved":false,"context_lines":[{"line_number":94,"context_line":""},{"line_number":95,"context_line":"    def _process_queue_loop(self):"},{"line_number":96,"context_line":"        LOG.debug(\"Starting ipconntrack _process_queue_loop()\")"},{"line_number":97,"context_line":"        pool \u003d eventlet.GreenPool(size\u003d8)"},{"line_number":98,"context_line":"        while True:"},{"line_number":99,"context_line":"            pool.spawn_n(self._process_queue)"},{"line_number":100,"context_line":""}],"source_content_type":"text/x-python","patch_set":3,"id":"3fa0c359_77d0c0ca","line":97,"range":{"start_line":97,"start_character":34,"end_line":97,"end_character":40},"in_reply_to":"3fa0c359_9cbc9abc","updated":"2018-02-05 20:36:16.000000000","message":"I chose 8 since it was the same as the l3-agent queue processing code in neutron/agent/l3/agent.py - that\u0027s where this was copied from basically.  But yes, this shouldn\u0027t be hard-coded here, should use conntrack_workers from above.  Luckily my testing was all using 8 :)","commit_id":"3294e923e7449d207b973de78716965a1a4770de"}],"neutron/conf/agent/securitygroups_rpc.py":[{"author":{"_account_id":11975,"name":"Slawek Kaplonski","email":"skaplons@redhat.com","username":"slaweq"},"change_message_id":"12a7bb1115c345b7b6545299aede38815cdceddd","unresolved":false,"context_lines":[{"line_number":44,"context_line":"        help\u003d_(\u0027Number of conntrack workers used to speed-up the removal of \u0027"},{"line_number":45,"context_line":"               \u0027conntrack entries when security groups are updated. \u0027"},{"line_number":46,"context_line":"               \u0027Setting to zero will disable workers and do all updates \u0027"},{"line_number":47,"context_line":"               \u0027in the calling thread.\u0027))"},{"line_number":48,"context_line":"]"},{"line_number":49,"context_line":""},{"line_number":50,"context_line":""}],"source_content_type":"text/x-python","patch_set":3,"id":"3fa0c359_dc722212","line":47,"updated":"2018-02-05 11:15:09.000000000","message":"What about introducing one another option like \"use_conntrack_parallel\" \u003d {True,False} and here if number of workers will be 0 then determine it automatically , e.g. set same as number of cores or something like that","commit_id":"3294e923e7449d207b973de78716965a1a4770de"},{"author":{"_account_id":1131,"name":"Brian Haley","email":"haleyb.dev@gmail.com","username":"brian-haley"},"change_message_id":"8c09195e72cf1dd4f7c02cd11620df4198de75a4","unresolved":false,"context_lines":[{"line_number":44,"context_line":"        help\u003d_(\u0027Number of conntrack workers used to speed-up the removal of \u0027"},{"line_number":45,"context_line":"               \u0027conntrack entries when security groups are updated. \u0027"},{"line_number":46,"context_line":"               \u0027Setting to zero will disable workers and do all updates \u0027"},{"line_number":47,"context_line":"               \u0027in the calling thread.\u0027))"},{"line_number":48,"context_line":"]"},{"line_number":49,"context_line":""},{"line_number":50,"context_line":""}],"source_content_type":"text/x-python","patch_set":3,"id":"3fa0c359_846c3109","line":47,"in_reply_to":"3fa0c359_dc722212","updated":"2018-02-05 20:36:16.000000000","message":"I\u0027d like to keep a way to be able to disable this optimization, which is zero.\n\nI guess I could make this (1+ host.cpu_count() // 2) like is done for other things like metadata workers, I just might have to do some performance checking to make sure that using one other thread is good enough in the base case.","commit_id":"3294e923e7449d207b973de78716965a1a4770de"}]}
