)]}'
{"oslo_messaging/_drivers/impl_rabbit.py":[{"author":{"_account_id":8601,"name":"Dmitriy Ukhlov","email":"dukhlov@mirantis.com","username":"dukhlov"},"change_message_id":"aa96f058df53c2f3b7b10edcd2ac372a57ceb0be","unresolved":false,"context_lines":[{"line_number":139,"context_line":"                help\u003d\u0027Use HA queues in RabbitMQ (x-ha-policy: all). \u0027"},{"line_number":140,"context_line":"                     \u0027If you change this option, you must wipe the \u0027"},{"line_number":141,"context_line":"                     \u0027RabbitMQ database.\u0027),"},{"line_number":142,"context_line":"    cfg.IntOpt(\u0027rabbit_qos_prefetch\u0027,"},{"line_number":143,"context_line":"               default\u003d0,"},{"line_number":144,"context_line":"               help\u003d\u0027Specifies the number of messages to prefetch. Setting to \u0027"},{"line_number":145,"context_line":"                    \u0027zero allows unlimited messages.\u0027),"}],"source_content_type":"text/x-python","patch_set":1,"id":"fa69d971_838bdf1d","line":142,"range":{"start_line":142,"start_character":16,"end_line":142,"end_character":35},"updated":"2016-01-08 20:44:59.000000000","message":"I see possible problem when we set rabbit_qos_prefetch to 10 for examle and then call listener.poll(prefetch_size\u003d100)\nIn this case method will never receive 100 unacked messages and will wait forever (or in scope of timeout what is undesired too)","commit_id":"1e936a6f99ffdd0b0b10b114f40c62187fdb2e71"},{"author":{"_account_id":8601,"name":"Dmitriy Ukhlov","email":"dukhlov@mirantis.com","username":"dukhlov"},"change_message_id":"aa96f058df53c2f3b7b10edcd2ac372a57ceb0be","unresolved":false,"context_lines":[{"line_number":143,"context_line":"               default\u003d0,"},{"line_number":144,"context_line":"               help\u003d\u0027Specifies the number of messages to prefetch. Setting to \u0027"},{"line_number":145,"context_line":"                    \u0027zero allows unlimited messages.\u0027),"},{"line_number":146,"context_line":"    cfg.BoolOpt(\u0027rabbit_qos_global\u0027,"},{"line_number":147,"context_line":"                default\u003dFalse,"},{"line_number":148,"context_line":"                help\u003d\u0027RabbitMQ takes global\u003dfalse to mean that the QoS \u0027"},{"line_number":149,"context_line":"                     \u0027settings should apply per-consumer (for new consumers \u0027"}],"source_content_type":"text/x-python","patch_set":1,"id":"fa69d971_63d34b39","line":146,"range":{"start_line":146,"start_character":17,"end_line":146,"end_character":34},"updated":"2016-01-08 20:44:59.000000000","message":"I dont think that we need rabbit_qos_global option. \u0027rabbit_*\u0027 options are independent of oslo.messaging patterns and configure just transport. Using these you could tune delivery mode and tune performace. Setting rabbit_qos_global to True will just slow down message consumption and will not provide any benefits. Probably it is resonable to make some oslo.messaging specific option - prefetch for rpc_endpoint for example (not for queue), but it\u0027s another usecase","commit_id":"1e936a6f99ffdd0b0b10b114f40c62187fdb2e71"}]}