)]}'
{"priorities/train-priorities.rst":[{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"747d67b6bd267e35e1ca712c839d6ba4aee84166","unresolved":false,"context_lines":[{"line_number":12,"context_line":""},{"line_number":13,"context_line":"#. **Improve/expand what we track in Placement, so we can ask Placement better"},{"line_number":14,"context_line":"   questions, so deployments - especially large-scale deployments - can schedule"},{"line_number":15,"context_line":"   more efficiently and effectively.** This encompasses efforts such as:"},{"line_number":16,"context_line":""},{"line_number":17,"context_line":"   * NUMA structure modeling and affinity"},{"line_number":18,"context_line":"   * Improved tracking of shared and dedicated logical processors"}],"source_content_type":"text/x-rst","patch_set":1,"id":"dfbec78f_745f3e72","line":15,"updated":"2019-05-07 14:13:00.000000000","message":"The above isn\u0027t about usecases. While I understand it and agree with the below, I think we should explain *why* we could have this as a theme.","commit_id":"6638d10af330660a28fd5aa663e35d21de33b632"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"5da88cae77cd2996fe00a2fbaf7a2782adebc623","unresolved":false,"context_lines":[{"line_number":12,"context_line":""},{"line_number":13,"context_line":"#. **Improve/expand what we track in Placement, so we can ask Placement better"},{"line_number":14,"context_line":"   questions, so deployments - especially large-scale deployments - can schedule"},{"line_number":15,"context_line":"   more efficiently and effectively.** This encompasses efforts such as:"},{"line_number":16,"context_line":""},{"line_number":17,"context_line":"   * NUMA structure modeling and affinity"},{"line_number":18,"context_line":"   * Improved tracking of shared and dedicated logical processors"}],"source_content_type":"text/x-rst","patch_set":1,"id":"dfbec78f_772500ab","line":15,"in_reply_to":"dfbec78f_745f3e72","updated":"2019-05-07 14:23:22.000000000","message":"TBC, I don\u0027t disagree with having the below be a theme for Train. My concerns are just making sure we can explain to our users and operators which usecases we would be prioriting for Train with Placement.\n\nAn alternative proposal could be : \n- Improve/expand Nova scheduling efficience and scalability for all deployments, with a particular focus on large-scale ones.\n\nI\u0027m maybe just nitpicking but I\u0027d just remove mention of Placement in the first sentence.","commit_id":"6638d10af330660a28fd5aa663e35d21de33b632"},{"author":{"_account_id":14070,"name":"Eric Fried","email":"openstack@fried.cc","username":"efried"},"change_message_id":"9e628cf620beb8fc98f951bf00d0312c8645fc2d","unresolved":false,"context_lines":[{"line_number":12,"context_line":""},{"line_number":13,"context_line":"#. **Improve/expand what we track in Placement, so we can ask Placement better"},{"line_number":14,"context_line":"   questions, so deployments - especially large-scale deployments - can schedule"},{"line_number":15,"context_line":"   more efficiently and effectively.** This encompasses efforts such as:"},{"line_number":16,"context_line":""},{"line_number":17,"context_line":"   * NUMA structure modeling and affinity"},{"line_number":18,"context_line":"   * Improved tracking of shared and dedicated logical processors"}],"source_content_type":"text/x-rst","patch_set":1,"id":"dfbec78f_b6d8357f","line":15,"in_reply_to":"dfbec78f_772500ab","updated":"2019-05-09 14:08:52.000000000","message":"I don\u0027t agree the Stein goal [1] was particularly strict about being exclusively user-facing.\n\nBut I reworded this so the bold part of the goal was more general as you suggested, and then afterward explained how that is to be done (i.e. via placement-isms). See what you think.\n\n[1] In particular, \"Compute nodes capable to upgrade and exist with nested resource providers for multiple GPU types\" had/has no user-facing impact.","commit_id":"6638d10af330660a28fd5aa663e35d21de33b632"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"274158e594f2eca1081e1c5abb37345c0a758432","unresolved":false,"context_lines":[{"line_number":12,"context_line":""},{"line_number":13,"context_line":"#. **Improve/expand what we track in Placement, so we can ask Placement better"},{"line_number":14,"context_line":"   questions, so deployments - especially large-scale deployments - can schedule"},{"line_number":15,"context_line":"   more efficiently and effectively.** This encompasses efforts such as:"},{"line_number":16,"context_line":""},{"line_number":17,"context_line":"   * NUMA structure modeling and affinity"},{"line_number":18,"context_line":"   * Improved tracking of shared and dedicated logical processors"}],"source_content_type":"text/x-rst","patch_set":1,"id":"dfbec78f_843a8931","line":15,"in_reply_to":"dfbec78f_b6d8357f","updated":"2019-05-10 07:38:25.000000000","message":"Better resource modelling / management leads to better scheduling performance and that has significant deployer facing value for my employer.","commit_id":"6638d10af330660a28fd5aa663e35d21de33b632"},{"author":{"_account_id":14070,"name":"Eric Fried","email":"openstack@fried.cc","username":"efried"},"change_message_id":"87da596697624853d9781de6df8c751b6fa0d65e","unresolved":false,"context_lines":[{"line_number":20,"context_line":"   * Supporting server group affinity and anti-affinity"},{"line_number":21,"context_line":"   * Trait filters for driver capabilities, image types, and more"},{"line_number":22,"context_line":"   * Forbidden trait and aggregate filters to isolate \"special\" hosts, avoid"},{"line_number":23,"context_line":"     disabled nodes, etc."},{"line_number":24,"context_line":""},{"line_number":25,"context_line":"   .. note:: Much of the above work has dependencies on efforts in Placement."},{"line_number":26,"context_line":"             The Placement team\u0027s cycle priorities are aligned accordingly."}],"source_content_type":"text/x-rst","patch_set":1,"id":"dfbec78f_2b77413d","line":23,"updated":"2019-05-04 23:32:55.000000000","message":"Should we make some of these links to their respective blueprints/specs?","commit_id":"6638d10af330660a28fd5aa663e35d21de33b632"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"747d67b6bd267e35e1ca712c839d6ba4aee84166","unresolved":false,"context_lines":[{"line_number":20,"context_line":"   * Supporting server group affinity and anti-affinity"},{"line_number":21,"context_line":"   * Trait filters for driver capabilities, image types, and more"},{"line_number":22,"context_line":"   * Forbidden trait and aggregate filters to isolate \"special\" hosts, avoid"},{"line_number":23,"context_line":"     disabled nodes, etc."},{"line_number":24,"context_line":""},{"line_number":25,"context_line":"   .. note:: Much of the above work has dependencies on efforts in Placement."},{"line_number":26,"context_line":"             The Placement team\u0027s cycle priorities are aligned accordingly."}],"source_content_type":"text/x-rst","patch_set":1,"id":"dfbec78f_b4ca9617","line":23,"in_reply_to":"dfbec78f_2b77413d","updated":"2019-05-07 14:13:00.000000000","message":"I don\u0027t think so. Those themes are about usecases, not specifications.","commit_id":"6638d10af330660a28fd5aa663e35d21de33b632"},{"author":{"_account_id":14070,"name":"Eric Fried","email":"openstack@fried.cc","username":"efried"},"change_message_id":"9e628cf620beb8fc98f951bf00d0312c8645fc2d","unresolved":false,"context_lines":[{"line_number":20,"context_line":"   * Supporting server group affinity and anti-affinity"},{"line_number":21,"context_line":"   * Trait filters for driver capabilities, image types, and more"},{"line_number":22,"context_line":"   * Forbidden trait and aggregate filters to isolate \"special\" hosts, avoid"},{"line_number":23,"context_line":"     disabled nodes, etc."},{"line_number":24,"context_line":""},{"line_number":25,"context_line":"   .. note:: Much of the above work has dependencies on efforts in Placement."},{"line_number":26,"context_line":"             The Placement team\u0027s cycle priorities are aligned accordingly."}],"source_content_type":"text/x-rst","patch_set":1,"id":"dfbec78f_16d34162","line":23,"in_reply_to":"dfbec78f_b4ca9617","updated":"2019-05-09 14:08:52.000000000","message":"ack","commit_id":"6638d10af330660a28fd5aa663e35d21de33b632"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"747d67b6bd267e35e1ca712c839d6ba4aee84166","unresolved":false,"context_lines":[{"line_number":26,"context_line":"             The Placement team\u0027s cycle priorities are aligned accordingly."},{"line_number":27,"context_line":""},{"line_number":28,"context_line":"#. **Enable requesting an instance with one or more accelerators either"},{"line_number":29,"context_line":"   preprogrammed or dynamically programmed.**"},{"line_number":30,"context_line":""},{"line_number":31,"context_line":"   .. note:: This is a cross-project theme with Cyborg. The Cyborg team\u0027s cycle"},{"line_number":32,"context_line":"             priorities are aligned accordingly."}],"source_content_type":"text/x-rst","patch_set":1,"id":"dfbec78f_54831acd","line":29,"updated":"2019-05-07 14:13:00.000000000","message":"I like the usecase, as it\u0027s not Cyborg-specific.\nWhat we should then agree is about accepting changes about accelerators (like virtual GPUs ;-) ) as related priorities for this theme, and not only Cyborg.","commit_id":"6638d10af330660a28fd5aa663e35d21de33b632"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"5da88cae77cd2996fe00a2fbaf7a2782adebc623","unresolved":false,"context_lines":[{"line_number":30,"context_line":""},{"line_number":31,"context_line":"   .. note:: This is a cross-project theme with Cyborg. The Cyborg team\u0027s cycle"},{"line_number":32,"context_line":"             priorities are aligned accordingly."},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"#. **We want our documentation to be valid, easily referenced and generally"},{"line_number":35,"context_line":"   suitable for purpose.** We\u0027re building on a strong foundation. Three"},{"line_number":36,"context_line":"   objectives:"}],"source_content_type":"text/x-rst","patch_set":1,"id":"dfbec78f_974d1474","line":33,"updated":"2019-05-07 14:23:22.000000000","message":"I also think we should more explain which specific things to enhance here (and just not say \"heh, Cyborg\")","commit_id":"6638d10af330660a28fd5aa663e35d21de33b632"},{"author":{"_account_id":14070,"name":"Eric Fried","email":"openstack@fried.cc","username":"efried"},"change_message_id":"9e628cf620beb8fc98f951bf00d0312c8645fc2d","unresolved":false,"context_lines":[{"line_number":30,"context_line":""},{"line_number":31,"context_line":"   .. note:: This is a cross-project theme with Cyborg. The Cyborg team\u0027s cycle"},{"line_number":32,"context_line":"             priorities are aligned accordingly."},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"#. **We want our documentation to be valid, easily referenced and generally"},{"line_number":35,"context_line":"   suitable for purpose.** We\u0027re building on a strong foundation. Three"},{"line_number":36,"context_line":"   objectives:"}],"source_content_type":"text/x-rst","patch_set":1,"id":"dfbec78f_b6c11583","line":33,"in_reply_to":"dfbec78f_974d1474","updated":"2019-05-09 14:08:52.000000000","message":"Done, I think. Do you have more in mind than VGPUs?","commit_id":"6638d10af330660a28fd5aa663e35d21de33b632"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"747d67b6bd267e35e1ca712c839d6ba4aee84166","unresolved":false,"context_lines":[{"line_number":44,"context_line":"      should be able to search Google for e.g. \"attaching a PCI device\" and"},{"line_number":45,"context_line":"      find the guide that details what it is, how to enable it and how to use"},{"line_number":46,"context_line":"      it. See https://docs.djangoproject.com/en/2.2/topics/migrations/ for a"},{"line_number":47,"context_line":"      non-OpenStack variant."},{"line_number":48,"context_line":""},{"line_number":49,"context_line":".. _`console docs`: https://docs.openstack.org/nova/latest/admin/remote-console-access.html#novnc-based-vnc-console"}],"source_content_type":"text/x-rst","patch_set":1,"id":"dfbec78f_74d41ebe","line":47,"updated":"2019-05-07 14:13:00.000000000","message":"++","commit_id":"6638d10af330660a28fd5aa663e35d21de33b632"}]}