)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"065ad41a308335c225bc428abe91038b0daa5a1d","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"50d48f33_18686e32","updated":"2023-12-15 09:33:09.000000000","message":"\u003e At a minimum, this needs to be configurable in the same way that periodic_max_workers is, then.\n\nThere is a configuration option, isn\u0027t it enough?\n\nI don\u0027t mind the opposite feature, but I don\u0027t have a valid case for it, so I\u0027d prefer someone who does to introduce it.","commit_id":"adec0f6f01a1733bbdbe8e1f962138e5c7363e90"},{"author":{"_account_id":10342,"name":"Jay Faulkner","display_name":"JayF","email":"jay@jvf.cc","username":"JayF","status":"youtube.com/@oss-gr / podcast.gr-oss.io"},"change_message_id":"f2baba82bc23c568d86fdaaf9112b82f26f70daf","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"6b917e60_a54fa4eb","updated":"2023-12-14 21:45:52.000000000","message":"At a minimum, this needs to be configurable in the same way that periodic_max_workers is, then.","commit_id":"adec0f6f01a1733bbdbe8e1f962138e5c7363e90"},{"author":{"_account_id":10342,"name":"Jay Faulkner","display_name":"JayF","email":"jay@jvf.cc","username":"JayF","status":"youtube.com/@oss-gr / podcast.gr-oss.io"},"change_message_id":"b45a89560ce5018fe0da19e334976497dbb65b42","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"f04da745_107e3049","updated":"2023-12-13 21:47:00.000000000","message":"Is there other approaches we can take? Using a separate pool of workers, like we do for periodic tasks?\n\nI don\u0027t hate the idea of this, but it seems like drawing the line at heartbeats only is weird, and I could see the opposite case -- where an operator would want all other API calls to fail in service of making provisioning activities (driven by heartbeats) succeed.","commit_id":"adec0f6f01a1733bbdbe8e1f962138e5c7363e90"},{"author":{"_account_id":24828,"name":"Kaifeng Wang","email":"kaifeng.w@gmail.com","username":"wangkf"},"change_message_id":"1c8a396aaff2bd4a435ed83258a0132b70de9150","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"da7e3ae1_45e3d34c","updated":"2024-01-02 05:48:25.000000000","message":"Kind of like QoS, it would be straightforward if the executor supports prioritized scheduling. In the large scale scenario, I guess many configuration options need to be increased accordingly, and scaling issues can\u0027t be addressed by the pool size alone.","commit_id":"adec0f6f01a1733bbdbe8e1f962138e5c7363e90"},{"author":{"_account_id":10342,"name":"Jay Faulkner","display_name":"JayF","email":"jay@jvf.cc","username":"JayF","status":"youtube.com/@oss-gr / podcast.gr-oss.io"},"change_message_id":"7d3bcef0805cebb5749ecbac3127b1d409d20834","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"ca7b63e3_cd6f2329","updated":"2023-12-18 15:37:59.000000000","message":"This is going to add some operational complexity that I\u0027m not super keen on, but it also solves real world problems. I don\u0027t like this solution but feel it\u0027s not right for me to -1 it without a viable alternative.","commit_id":"adec0f6f01a1733bbdbe8e1f962138e5c7363e90"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"d7662234e6e0b887d5f07d4fb6a7ecab7dcf88b2","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"d9db54fe_86d5f5ad","updated":"2024-02-29 00:48:23.000000000","message":"recheck ironic-standalone is being moved to non-voting.","commit_id":"adec0f6f01a1733bbdbe8e1f962138e5c7363e90"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"bc829f25ba909563bec454c7498c899288ab663e","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"94f43221_d04a6308","updated":"2024-02-29 04:48:09.000000000","message":"recheck ironic-standalone is now non voting","commit_id":"adec0f6f01a1733bbdbe8e1f962138e5c7363e90"},{"author":{"_account_id":10342,"name":"Jay Faulkner","display_name":"JayF","email":"jay@jvf.cc","username":"JayF","status":"youtube.com/@oss-gr / podcast.gr-oss.io"},"change_message_id":"a4e7808266285f223e394fdd2848d542ff90bbdc","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"b480fe2a_b09fba48","in_reply_to":"ea222db9_ab45c611","updated":"2023-12-14 21:47:37.000000000","message":"But it\u0027s not a binary choice: it\u0027s a sliding scale. How many heartbeats are you OK with choking out the API? And using a static percentage makes it weird at higher scale? I think there\u0027s room here for a configuration option, at a minimum, to give more control to the deployer.","commit_id":"adec0f6f01a1733bbdbe8e1f962138e5c7363e90"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"a7b23e68c58841b2aca86277250a299593f9ed03","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"ea222db9_ab45c611","in_reply_to":"f04da745_107e3049","updated":"2023-12-14 07:20:06.000000000","message":"I can hardly imagine someone saying \"yes, please, remove all my control over Ironic in favour of the internal business that may not converge any soon\". Keep in mind: heartbeats are trivial to retry (5-10 seconds after my recent patch). Failed heartbeats don\u0027t prevent deployments from proceeding eventually.\n\nBtw, periodic tasks do not use a separate pool of workers. It\u0027s same same pool (main pool in this patch).","commit_id":"adec0f6f01a1733bbdbe8e1f962138e5c7363e90"}]}