)]}'
{"specs/supported-messaging-drivers.rst":[{"author":{"_account_id":8770,"name":"Ken Giusti","email":"kgiusti@gmail.com","username":"kgiusti"},"change_message_id":"dd1efde51948f304f8f565ef6c34a290056b7dd8","unresolved":false,"context_lines":[{"line_number":11,"context_line":""},{"line_number":12,"context_line":"OpenStack has gravitated toward using RabbitMQ for message passing. There"},{"line_number":13,"context_line":"are numerous excellent alternatives available as backends, including"},{"line_number":14,"context_line":"QPID and 0mq, but only RabbitMQ has received attention direclty in the"},{"line_number":15,"context_line":"community at large. As a result, more and more users of these backends"},{"line_number":16,"context_line":"are switching to RabbitMQ, and this leaves QPID code lying around that"},{"line_number":17,"context_line":"is not well tested and not well supported by the community. Said code"}],"source_content_type":"text/x-rst","patch_set":1,"id":"5aa745d1_b6269ab8","line":14,"updated":"2015-04-16 16:07:07.000000000","message":"\"directly\"","commit_id":"ff19d9ba86cd18459bb912acdb9cedccc6076151"},{"author":{"_account_id":8770,"name":"Ken Giusti","email":"kgiusti@gmail.com","username":"kgiusti"},"change_message_id":"dd1efde51948f304f8f565ef6c34a290056b7dd8","unresolved":false,"context_lines":[{"line_number":53,"context_line":"* Must have integration testing including at least 3 popular oslo.messaging"},{"line_number":54,"context_line":"  dependents, preferrably at the minimum a devstack-gate job with Nova,"},{"line_number":55,"context_line":"  Cinder, and Neutron."},{"line_number":56,"context_line":""},{"line_number":57,"context_line":"* All testing above must be voting in the gate of oslo.messaging."},{"line_number":58,"context_line":""},{"line_number":59,"context_line":"Documentation"}],"source_content_type":"text/x-rst","patch_set":1,"id":"5aa745d1_96d5de44","line":56,"updated":"2015-04-16 16:07:07.000000000","message":"This presents a chicken-or-egg dilemma: there can\u0027t be CI tests until _after_ the driver is available in oslo.messaging (at least that\u0027s been my experience with amqp1.0).  Can we add some allowance for introducing the driver as experimental until the required CI tests are present?","commit_id":"ff19d9ba86cd18459bb912acdb9cedccc6076151"},{"author":{"_account_id":6488,"name":"Clint Byrum","email":"clint@fewbar.com","username":"clint-fewbar"},"change_message_id":"616934f3245adc14010a331d008c05b66ff26d78","unresolved":false,"context_lines":[{"line_number":53,"context_line":"* Must have integration testing including at least 3 popular oslo.messaging"},{"line_number":54,"context_line":"  dependents, preferrably at the minimum a devstack-gate job with Nova,"},{"line_number":55,"context_line":"  Cinder, and Neutron."},{"line_number":56,"context_line":""},{"line_number":57,"context_line":"* All testing above must be voting in the gate of oslo.messaging."},{"line_number":58,"context_line":""},{"line_number":59,"context_line":"Documentation"}],"source_content_type":"text/x-rst","patch_set":1,"id":"3ab451fb_823a5ecc","line":56,"in_reply_to":"5aa745d1_96d5de44","updated":"2015-04-16 20:59:22.000000000","message":"Great point, I\u0027ll add some concessions for new drivers.","commit_id":"ff19d9ba86cd18459bb912acdb9cedccc6076151"},{"author":{"_account_id":8770,"name":"Ken Giusti","email":"kgiusti@gmail.com","username":"kgiusti"},"change_message_id":"dd1efde51948f304f8f565ef6c34a290056b7dd8","unresolved":false,"context_lines":[{"line_number":55,"context_line":"  Cinder, and Neutron."},{"line_number":56,"context_line":""},{"line_number":57,"context_line":"* All testing above must be voting in the gate of oslo.messaging."},{"line_number":58,"context_line":""},{"line_number":59,"context_line":"Documentation"},{"line_number":60,"context_line":"~~~~~~~~~~~~~"},{"line_number":61,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"5aa745d1_96639eb1","line":58,"updated":"2015-04-16 16:07:07.000000000","message":"It would help greatly if there were links to resources explaining how to create CI tests.","commit_id":"ff19d9ba86cd18459bb912acdb9cedccc6076151"},{"author":{"_account_id":6488,"name":"Clint Byrum","email":"clint@fewbar.com","username":"clint-fewbar"},"change_message_id":"616934f3245adc14010a331d008c05b66ff26d78","unresolved":false,"context_lines":[{"line_number":55,"context_line":"  Cinder, and Neutron."},{"line_number":56,"context_line":""},{"line_number":57,"context_line":"* All testing above must be voting in the gate of oslo.messaging."},{"line_number":58,"context_line":""},{"line_number":59,"context_line":"Documentation"},{"line_number":60,"context_line":"~~~~~~~~~~~~~"},{"line_number":61,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"3ab451fb_423456dd","line":58,"in_reply_to":"5aa745d1_96639eb1","updated":"2015-04-16 20:59:22.000000000","message":"Another great point. Will do in upcoming revisions.","commit_id":"ff19d9ba86cd18459bb912acdb9cedccc6076151"},{"author":{"_account_id":1297,"name":"Joshua Harlow","email":"jxharlow@godaddy.com","username":"harlowja"},"change_message_id":"cf50005353e655f3b995ca2cce567299d009dfba","unresolved":false,"context_lines":[{"line_number":76,"context_line":"ship RabbitMQ drivers. This option would alienate users who have private"},{"line_number":77,"context_line":"drivers, and would also force users of the zmq driver who are trying to"},{"line_number":78,"context_line":"fix it to abandon those efforts and try to scale with RabbitMQ."},{"line_number":79,"context_line":""},{"line_number":80,"context_line":"Implementation"},{"line_number":81,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":82,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"5aa745d1_776f1093","line":79,"updated":"2015-04-15 21:31:24.000000000","message":"Or just have oslo.messaging more directly wrap kombu (aka have oslo.messaging provide some basic rpc patterns and configuration patterns via oslo.config), and let people use \u0027experimental\u0027 drivers via kombu but say that the rabbitmq one from kombu is the most well-known to work; this also helps the kombu community (and any users that also use kombu, which afaik is non-zero, since celery uses kombu, and taskflow does also..)\n\nIf https://github.com/celery/kombu/issues/430 ever happened this could shrink down those \u0027basic rpc patterns\u0027 down to zero (or nearly zero); letting openstack avoid focusing on messaging (and letting the kombu community do this for us instead); letting us focus on bigger \u0026 better things...","commit_id":"ff19d9ba86cd18459bb912acdb9cedccc6076151"},{"author":{"_account_id":2472,"name":"Doug Hellmann","email":"dhellmann@redhat.com","username":"doug-hellmann"},"change_message_id":"d286a4997fde5fda731d45ae9fe75205bda12490","unresolved":false,"context_lines":[{"line_number":76,"context_line":"ship RabbitMQ drivers. This option would alienate users who have private"},{"line_number":77,"context_line":"drivers, and would also force users of the zmq driver who are trying to"},{"line_number":78,"context_line":"fix it to abandon those efforts and try to scale with RabbitMQ."},{"line_number":79,"context_line":""},{"line_number":80,"context_line":"Implementation"},{"line_number":81,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":82,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"3ab451fb_c598706d","line":79,"in_reply_to":"3ab451fb_bfb3330a","updated":"2015-04-16 21:36:53.000000000","message":"I\u0027m not sure how practical it really is to suggest such a large change. We could consider it, but it would require a much more detailed spec, and would need to be considered separately from the incremental change proposed here.","commit_id":"ff19d9ba86cd18459bb912acdb9cedccc6076151"},{"author":{"_account_id":1297,"name":"Joshua Harlow","email":"jxharlow@godaddy.com","username":"harlowja"},"change_message_id":"00984b68e7854de54fb22d0a5757399440a14838","unresolved":false,"context_lines":[{"line_number":76,"context_line":"ship RabbitMQ drivers. This option would alienate users who have private"},{"line_number":77,"context_line":"drivers, and would also force users of the zmq driver who are trying to"},{"line_number":78,"context_line":"fix it to abandon those efforts and try to scale with RabbitMQ."},{"line_number":79,"context_line":""},{"line_number":80,"context_line":"Implementation"},{"line_number":81,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":82,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"3ab451fb_7648302e","line":79,"in_reply_to":"3ab451fb_c598706d","updated":"2015-04-16 23:13:53.000000000","message":"Fair enough.","commit_id":"ff19d9ba86cd18459bb912acdb9cedccc6076151"},{"author":{"_account_id":6159,"name":"Flavio Percoco Premoli","display_name":"flaper87","email":"flavio.percoco@flyrlabs.com","username":"flaper87"},"change_message_id":"c58b6d3e97f64dc6af5873bba7a7c4aa581ab477","unresolved":false,"context_lines":[{"line_number":76,"context_line":"ship RabbitMQ drivers. This option would alienate users who have private"},{"line_number":77,"context_line":"drivers, and would also force users of the zmq driver who are trying to"},{"line_number":78,"context_line":"fix it to abandon those efforts and try to scale with RabbitMQ."},{"line_number":79,"context_line":""},{"line_number":80,"context_line":"Implementation"},{"line_number":81,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":82,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"3ab451fb_bfb3330a","line":79,"in_reply_to":"5aa745d1_2acfbab1","updated":"2015-04-16 20:28:36.000000000","message":"I\u0027ve push access to kombu so I know pretty well its license and workflow. The reason I made my comment is not because I\u0027m willing to re-invent the wheel. There\u0027s more to those o.m drivers than just \"support for yet another broker\".\n\nFor instance, kombu\u0027s model depends a lot on the exchange/queue/binding model which was taken from AMQP \u003c 1.0 This makes implementation of drivers like proton or 0mq a bit limited on the set of features that could take advantage of.\n\nSo no, I don\u0027t think having a single driver based on kombu is the way to go and no, I don\u0027t think we should switch and hear people crying because zmq doesn\u0027t work anymore. Last time I checked, there were users (same users that stepped up to maintain the driver).\n\nMaking a cloud reliable is not only about making messages go through it but also providing the right support for the technologies we are telling users to us. I don\u0027t think any member of the oslo team is playing at \"lets re-invent the cool things that have already been implemented\". In fact, whenever that has happened, we\u0027ve promptly addressed it replacing the code with whatever made more sense (see oslo.cache, for example).","commit_id":"ff19d9ba86cd18459bb912acdb9cedccc6076151"},{"author":{"_account_id":1297,"name":"Joshua Harlow","email":"jxharlow@godaddy.com","username":"harlowja"},"change_message_id":"ff1fc230239b0df341d1375a3b66def7015909f0","unresolved":false,"context_lines":[{"line_number":76,"context_line":"ship RabbitMQ drivers. This option would alienate users who have private"},{"line_number":77,"context_line":"drivers, and would also force users of the zmq driver who are trying to"},{"line_number":78,"context_line":"fix it to abandon those efforts and try to scale with RabbitMQ."},{"line_number":79,"context_line":""},{"line_number":80,"context_line":"Implementation"},{"line_number":81,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":82,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"5aa745d1_2acfbab1","line":79,"in_reply_to":"5aa745d1_36428a3f","updated":"2015-04-16 16:02:47.000000000","message":"As far as kombu+zmq I\u0027d rather know if anyone is using that first before even thinking about a spec, from what I am starting to feel it appears the 0MQ stuff is in the same boat as QPID. In all honesty I\u0027d rather stop having oslo.* libraries containing experimental driver(s) and leave that experimentation to some other project (kombu?); after all (IMHO) it shouldn\u0027t be our expertise/area to experiment, but instead we should be focusing on making a damn reliable cloud....","commit_id":"ff19d9ba86cd18459bb912acdb9cedccc6076151"},{"author":{"_account_id":6159,"name":"Flavio Percoco Premoli","display_name":"flaper87","email":"flavio.percoco@flyrlabs.com","username":"flaper87"},"change_message_id":"ac05cdb2465a35c8cdd995a969556aa65254a379","unresolved":false,"context_lines":[{"line_number":76,"context_line":"ship RabbitMQ drivers. This option would alienate users who have private"},{"line_number":77,"context_line":"drivers, and would also force users of the zmq driver who are trying to"},{"line_number":78,"context_line":"fix it to abandon those efforts and try to scale with RabbitMQ."},{"line_number":79,"context_line":""},{"line_number":80,"context_line":"Implementation"},{"line_number":81,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":82,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"5aa745d1_aee4150d","line":79,"in_reply_to":"5aa745d1_776f1093","updated":"2015-04-16 07:34:09.000000000","message":"The only reason I don\u0027t like this is because it will take lots of flexibility and control from us and our users when it comes to adding support for other transports/drivers. Probably not a big deal but we should keep it in mind.\n\nI\u0027d be interested in reviewing a spec that proposes using just kombu for rabbit and zmq if you\u0027re willing to write one. :) I\u0027d be curious to know if o.m+kombu+zmq will actually work.","commit_id":"ff19d9ba86cd18459bb912acdb9cedccc6076151"},{"author":{"_account_id":1297,"name":"Joshua Harlow","email":"jxharlow@godaddy.com","username":"harlowja"},"change_message_id":"61ef2cec816e495aaad8ff746f2dc89694ae028b","unresolved":false,"context_lines":[{"line_number":76,"context_line":"ship RabbitMQ drivers. This option would alienate users who have private"},{"line_number":77,"context_line":"drivers, and would also force users of the zmq driver who are trying to"},{"line_number":78,"context_line":"fix it to abandon those efforts and try to scale with RabbitMQ."},{"line_number":79,"context_line":""},{"line_number":80,"context_line":"Implementation"},{"line_number":81,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":82,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"5aa745d1_36428a3f","line":79,"in_reply_to":"5aa745d1_aee4150d","updated":"2015-04-16 15:42:16.000000000","message":"Huh, kombu is an opensource project, last time I checked u can pretty easily work with (and even change!) those kinds of projects... IMHO it is awkward already at this current time/now that we don\u0027t have someone directly working with the kombu folks (aka a kombu \u0027core\u0027) anyway since we are already so reliant on it... All this starts to feel like an openstack version of NIH syndrome...","commit_id":"ff19d9ba86cd18459bb912acdb9cedccc6076151"},{"author":{"_account_id":1297,"name":"Joshua Harlow","email":"jxharlow@godaddy.com","username":"harlowja"},"change_message_id":"bc28455d6eb887d69d3e54825ccfe6f88280436e","unresolved":false,"context_lines":[{"line_number":18,"context_line":"will continue to be a burden, and should be removed if it is doing more"},{"line_number":19,"context_line":"harm than good."},{"line_number":20,"context_line":""},{"line_number":21,"context_line":"There is also anecdotal evidence that users are using the zmq driver"},{"line_number":22,"context_line":"to achieve high scale with fixes, but the zmq driver is not well"},{"line_number":23,"context_line":"tested either, which may actually be worse than not having it at all,"},{"line_number":24,"context_line":"as now anecdotes are passed from user to user but results will be very"}],"source_content_type":"text/x-rst","patch_set":3,"id":"dac4157f_befe4f69","line":21,"updated":"2015-04-21 15:40:12.000000000","message":"Nit: should we just stick to zmq or 0mq (instead of both)?","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":964,"name":"Anne Gentle","email":"annegentle@justwriteclick.com","username":"annegentle"},"change_message_id":"e2af2e44f5e952cc0915a847a69285f7525b3e9a","unresolved":false,"context_lines":[{"line_number":18,"context_line":"will continue to be a burden, and should be removed if it is doing more"},{"line_number":19,"context_line":"harm than good."},{"line_number":20,"context_line":""},{"line_number":21,"context_line":"There is also anecdotal evidence that users are using the zmq driver"},{"line_number":22,"context_line":"to achieve high scale with fixes, but the zmq driver is not well"},{"line_number":23,"context_line":"tested either, which may actually be worse than not having it at all,"},{"line_number":24,"context_line":"as now anecdotes are passed from user to user but results will be very"}],"source_content_type":"text/x-rst","patch_set":3,"id":"3afb71cf_e30829dc","line":21,"in_reply_to":"dac4157f_7d5c33ed","updated":"2015-06-10 12:40:38.000000000","message":"In a cleanup round I\u0027d suggest 0mq (ZeroMQ) on first use then 0mq throughout.","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":6488,"name":"Clint Byrum","email":"clint@fewbar.com","username":"clint-fewbar"},"change_message_id":"a2e36c140c5b7e3004a8a9090d79bdf906ed28fa","unresolved":false,"context_lines":[{"line_number":18,"context_line":"will continue to be a burden, and should be removed if it is doing more"},{"line_number":19,"context_line":"harm than good."},{"line_number":20,"context_line":""},{"line_number":21,"context_line":"There is also anecdotal evidence that users are using the zmq driver"},{"line_number":22,"context_line":"to achieve high scale with fixes, but the zmq driver is not well"},{"line_number":23,"context_line":"tested either, which may actually be worse than not having it at all,"},{"line_number":24,"context_line":"as now anecdotes are passed from user to user but results will be very"}],"source_content_type":"text/x-rst","patch_set":3,"id":"dac4157f_7d5c33ed","line":21,"in_reply_to":"dac4157f_befe4f69","updated":"2015-04-21 19:46:06.000000000","message":"Definitely worth establishing some taxonomy here at the top in the next round of edits.","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":1297,"name":"Joshua Harlow","email":"jxharlow@godaddy.com","username":"harlowja"},"change_message_id":"bc28455d6eb887d69d3e54825ccfe6f88280436e","unresolved":false,"context_lines":[{"line_number":30,"context_line":"RabbitMQ may not be sufficient for the entire community as the community"},{"line_number":31,"context_line":"grows. Pluggability is still something we should maintain, but we should"},{"line_number":32,"context_line":"have a very high standard for drivers that are shipped and documented"},{"line_number":33,"context_line":"as being supported."},{"line_number":34,"context_line":""},{"line_number":35,"context_line":"We will define a very clear policy as to the requirements for drivers"},{"line_number":36,"context_line":"to be carried in oslo.messaging and thus supported by the OpenStack"}],"source_content_type":"text/x-rst","patch_set":3,"id":"dac4157f_fe2ba7e9","line":33,"updated":"2015-04-21 15:40:12.000000000","message":"+1","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"0a5414695132beaa4b52c4979ef77fd843d3f7b1","unresolved":false,"context_lines":[{"line_number":39,"context_line":"channels to give users time to signal their needs. Deprecation will last"},{"line_number":40,"context_line":"for two release cycles before removing the code. We will also review and"},{"line_number":41,"context_line":"update documentation to annotate which drivers are supported and which"},{"line_number":42,"context_line":"are deprecated given these policies"},{"line_number":43,"context_line":""},{"line_number":44,"context_line":"Policy"},{"line_number":45,"context_line":"------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"dac4157f_8648d488","line":42,"updated":"2015-04-21 21:23:33.000000000","message":"just wondering the scenario where a driver is not up to standards and marked for \u0027deprecation\u0027... are we saying it can never be worked on again? and if not, what if the driver becomes compliant before 2 months? \n\nit seems slightly misleading to tag something deprecated if it can theorectically can come back.","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"f014d98149dfd31bac82e7cd28be9661e33e76be","unresolved":false,"context_lines":[{"line_number":39,"context_line":"channels to give users time to signal their needs. Deprecation will last"},{"line_number":40,"context_line":"for two release cycles before removing the code. We will also review and"},{"line_number":41,"context_line":"update documentation to annotate which drivers are supported and which"},{"line_number":42,"context_line":"are deprecated given these policies"},{"line_number":43,"context_line":""},{"line_number":44,"context_line":"Policy"},{"line_number":45,"context_line":"------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a0c5dc1_75dea66e","line":42,"in_reply_to":"9a0c5dc1_2a2b2d96","updated":"2015-05-13 16:31:51.000000000","message":"Using qpid as an example, devstack already supports qpid:\n\nhttp://git.openstack.org/cgit/openstack-dev/devstack/tree/lib/rpc_backend#n121\n\nSo we could have a job defined that configures devstack with qpid (via devstack-gate) and then run that on the experimental queue for oslo.messaging changes to make sure it\u0027s working, then make it non-voting in the check queue for o.m changes until the development team is happy with it being stable and makes it voting and gating.","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":6488,"name":"Clint Byrum","email":"clint@fewbar.com","username":"clint-fewbar"},"change_message_id":"d6ff7f8fb933fdf0154987cf0d1e46fe7b8f13c1","unresolved":false,"context_lines":[{"line_number":39,"context_line":"channels to give users time to signal their needs. Deprecation will last"},{"line_number":40,"context_line":"for two release cycles before removing the code. We will also review and"},{"line_number":41,"context_line":"update documentation to annotate which drivers are supported and which"},{"line_number":42,"context_line":"are deprecated given these policies"},{"line_number":43,"context_line":""},{"line_number":44,"context_line":"Policy"},{"line_number":45,"context_line":"------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"dac4157f_f19cecc1","line":42,"in_reply_to":"dac4157f_8648d488","updated":"2015-04-21 21:30:18.000000000","message":"Deprecation is not a death sentence. It is a policy tool to prevent investment in high-risk options. Of course it can be fixed and undeprecated. We can also re-introduce a driver that is removed and then fixed elsewhere.","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"05aafa91abd88ef32ee0511099e6d09b4e847043","unresolved":false,"context_lines":[{"line_number":39,"context_line":"channels to give users time to signal their needs. Deprecation will last"},{"line_number":40,"context_line":"for two release cycles before removing the code. We will also review and"},{"line_number":41,"context_line":"update documentation to annotate which drivers are supported and which"},{"line_number":42,"context_line":"are deprecated given these policies"},{"line_number":43,"context_line":""},{"line_number":44,"context_line":"Policy"},{"line_number":45,"context_line":"------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a0c5dc1_2a2b2d96","line":42,"in_reply_to":"dac4157f_f19cecc1","updated":"2015-05-13 16:29:10.000000000","message":"Similar to drivers in other projects, e.g. the docker virt driver in nova was in nova until icehouse when it didn\u0027t meet the QA requirements and was moved to stackforge, now it\u0027s been worked on there and is trying to get back into nova proper.","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"05aafa91abd88ef32ee0511099e6d09b4e847043","unresolved":false,"context_lines":[{"line_number":51,"context_line":"  reported by coverage report. Unit tests must be run for all versions"},{"line_number":52,"context_line":"  of python oslo.messaging currently gates on."},{"line_number":53,"context_line":""},{"line_number":54,"context_line":"* Must have integration testing including at least 3 popular oslo.messaging"},{"line_number":55,"context_line":"  dependents, preferrably at the minimum a devstack-gate job with Nova,"},{"line_number":56,"context_line":"  Cinder, and Neutron."},{"line_number":57,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a0c5dc1_95235aa7","line":54,"updated":"2015-05-13 16:29:10.000000000","message":"Could we just have, for example, a devstack job that configures qpid as the backend rather than rabbit, and runs devstack + tempest against oslo.messaging changes?  That would hit keystone, nova, cinder, glance, neutron/nova-network, etc.  We\u0027re doing that sort of thing with the ceph job for shared storage in nova.","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":8770,"name":"Ken Giusti","email":"kgiusti@gmail.com","username":"kgiusti"},"change_message_id":"fff6ca4640f896385ac375384824918e75168ca9","unresolved":false,"context_lines":[{"line_number":51,"context_line":"  reported by coverage report. Unit tests must be run for all versions"},{"line_number":52,"context_line":"  of python oslo.messaging currently gates on."},{"line_number":53,"context_line":""},{"line_number":54,"context_line":"* Must have integration testing including at least 3 popular oslo.messaging"},{"line_number":55,"context_line":"  dependents, preferrably at the minimum a devstack-gate job with Nova,"},{"line_number":56,"context_line":"  Cinder, and Neutron."},{"line_number":57,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a0c5dc1_14bd2368","line":54,"in_reply_to":"9a0c5dc1_95235aa7","updated":"2015-05-13 21:17:10.000000000","message":"+1","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":964,"name":"Anne Gentle","email":"annegentle@justwriteclick.com","username":"annegentle"},"change_message_id":"e2af2e44f5e952cc0915a847a69285f7525b3e9a","unresolved":false,"context_lines":[{"line_number":61,"context_line":"~~~~~~~~~~~~~"},{"line_number":62,"context_line":""},{"line_number":63,"context_line":"* Must have a reasonable amount of documentation including documentation"},{"line_number":64,"context_line":"  in the official OpenStack deployment guide."},{"line_number":65,"context_line":""},{"line_number":66,"context_line":"Support"},{"line_number":67,"context_line":"~~~~~~~"}],"source_content_type":"text/x-rst","patch_set":3,"id":"3afb71cf_0350e5d2","line":64,"updated":"2015-06-10 12:40:38.000000000","message":"There isn\u0027t really a deployment guide, but you could revise to \"official OpenStack Install Guides\" (since there are multiple guides per distro).","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":8770,"name":"Ken Giusti","email":"kgiusti@gmail.com","username":"kgiusti"},"change_message_id":"fff6ca4640f896385ac375384824918e75168ca9","unresolved":false,"context_lines":[{"line_number":68,"context_line":""},{"line_number":69,"context_line":"* Must have at least two individuals from the community commited to"},{"line_number":70,"context_line":"  triaging and fixing bugs, and responding to test failures in a timely"},{"line_number":71,"context_line":"  manner."},{"line_number":72,"context_line":""},{"line_number":73,"context_line":"Prospective Drivers"},{"line_number":74,"context_line":"~~~~~~~~~~~~~~~~~~~"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a0c5dc1_54303b01","line":71,"updated":"2015-05-13 21:17:10.000000000","message":"Sorry to admit my own cluelessness here, but is there a mechanism for notifying these individuals directly when specific test failure occurs?","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"47f06060efcbf7af39798d2638227191b45acaad","unresolved":false,"context_lines":[{"line_number":68,"context_line":""},{"line_number":69,"context_line":"* Must have at least two individuals from the community commited to"},{"line_number":70,"context_line":"  triaging and fixing bugs, and responding to test failures in a timely"},{"line_number":71,"context_line":"  manner."},{"line_number":72,"context_line":""},{"line_number":73,"context_line":"Prospective Drivers"},{"line_number":74,"context_line":"~~~~~~~~~~~~~~~~~~~"}],"source_content_type":"text/x-rst","patch_set":3,"id":"7a016987_190d9b3e","line":71,"in_reply_to":"7a016987_fd82bdc7","updated":"2015-05-14 19:01:10.000000000","message":"For the ceph job we just have a guy, jbernard, and if that job goes south we flame him up in IRC.  For the cells job, that is now alaski because he works for Rackspace and they added cells (and run with them).  These are examples.","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":6159,"name":"Flavio Percoco Premoli","display_name":"flaper87","email":"flavio.percoco@flyrlabs.com","username":"flaper87"},"change_message_id":"379c9c751ce28c4826c96d60398cf1e19087390f","unresolved":false,"context_lines":[{"line_number":68,"context_line":""},{"line_number":69,"context_line":"* Must have at least two individuals from the community commited to"},{"line_number":70,"context_line":"  triaging and fixing bugs, and responding to test failures in a timely"},{"line_number":71,"context_line":"  manner."},{"line_number":72,"context_line":""},{"line_number":73,"context_line":"Prospective Drivers"},{"line_number":74,"context_line":"~~~~~~~~~~~~~~~~~~~"}],"source_content_type":"text/x-rst","patch_set":3,"id":"7a016987_fd82bdc7","line":71,"in_reply_to":"9a0c5dc1_54303b01","updated":"2015-05-14 12:19:00.000000000","message":"You can watch patches that modify a specific code. As far as the gate failures go, you\u0027ll need to keep an eye on oslo.messaging. (Unless there\u0027s another trick I don\u0027t know)","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":6159,"name":"Flavio Percoco Premoli","display_name":"flaper87","email":"flavio.percoco@flyrlabs.com","username":"flaper87"},"change_message_id":"379c9c751ce28c4826c96d60398cf1e19087390f","unresolved":false,"context_lines":[{"line_number":77,"context_line":"  meet them will be given one full release cycle, or 6 months, whichever"},{"line_number":78,"context_line":"  is longer, to comply before being marked for deprecation. Their use,"},{"line_number":79,"context_line":"  however, will not be supported by the community. This will prevent a"},{"line_number":80,"context_line":"  chicken and egg problem for new drivers."},{"line_number":81,"context_line":""},{"line_number":82,"context_line":"Alternatives"},{"line_number":83,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"7a016987_ddd5c1b9","line":80,"updated":"2015-05-14 12:19:00.000000000","message":"We should also add warnings to drivers that don\u0027t meet these requirements as a way to communicate this to deployers. (Although very few ppl read logs)","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"47f06060efcbf7af39798d2638227191b45acaad","unresolved":false,"context_lines":[{"line_number":77,"context_line":"  meet them will be given one full release cycle, or 6 months, whichever"},{"line_number":78,"context_line":"  is longer, to comply before being marked for deprecation. Their use,"},{"line_number":79,"context_line":"  however, will not be supported by the community. This will prevent a"},{"line_number":80,"context_line":"  chicken and egg problem for new drivers."},{"line_number":81,"context_line":""},{"line_number":82,"context_line":"Alternatives"},{"line_number":83,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"7a016987_442566a9","line":80,"in_reply_to":"7a016987_ddd5c1b9","updated":"2015-05-14 19:01:10.000000000","message":"+1 for deprecation/experimental/untested warnings in the logs, nova has this on it\u0027s untested configurations for the most part (tooz, cells, libvirt + non-x86_64 architectures, lxc, etc).  It\u0027s less about people reading the logs and more about us covering our ass later when someone complains that the thing doesn\u0027t work and wasn\u0027t tested, we can say, yeah we know, it\u0027s in the logs right?","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"},{"author":{"_account_id":964,"name":"Anne Gentle","email":"annegentle@justwriteclick.com","username":"annegentle"},"change_message_id":"e2af2e44f5e952cc0915a847a69285f7525b3e9a","unresolved":false,"context_lines":[{"line_number":102,"context_line":"Work Items"},{"line_number":103,"context_line":"----------"},{"line_number":104,"context_line":""},{"line_number":105,"context_line":"- Record policy in oslo developer documentation"},{"line_number":106,"context_line":"- Announce policy to mailing lists"},{"line_number":107,"context_line":"- Mark non-compliant drivers as deprecated"},{"line_number":108,"context_line":"- Update configuration guides to note non-supported drivers"}],"source_content_type":"text/x-rst","patch_set":3,"id":"3afb71cf_a325915c","line":105,"updated":"2015-06-10 12:40:38.000000000","message":"Do we also need the policy pointer in ops and install docs?","commit_id":"fa30228a7152dd98fd1908714ce1576fd268c9a4"}]}