)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":35429,"name":"Biser Milanov","email":"biser.milanov@storpool.com","username":"sp-bmilanov"},"change_message_id":"475380ef59f153d25f6b537f2d35df21a35e5b76","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"6526bcfb_6913c0b1","updated":"2024-05-08 15:15:58.000000000","message":"-1 as it probably needs a blueprint and community discussions","commit_id":"b20e730cca63e388cf661585d4c1c18f86f8f99c"},{"author":{"_account_id":37484,"name":"Pavel Mracek","email":"mrak@gryg.net"},"change_message_id":"962d59adb16a3c40ed8177613b5521629381a8b0","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"98ae388b_61d47ebf","updated":"2024-11-15 12:30:41.000000000","message":"By the way, this is a very important feature. Without assigning io operations to the iothreads, the vCPU will continue to freeze during heavy and slow IO operations and then we can see messages like this in VMs dmesg:\n\n[Mon Feb 20 00:11:10 2023] watchdog: BUG: soft lockup - CPU#0 stuck for 347s! [kworker/u32:0:2615217]\n\nSo definitely thumbs up that someone is trying to implement it officially!","commit_id":"b20e730cca63e388cf661585d4c1c18f86f8f99c"},{"author":{"_account_id":37484,"name":"Pavel Mracek","email":"mrak@gryg.net"},"change_message_id":"c20cfdeb7a8885ba73edef43141beaa97307620f","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"6db4826a_071fda6c","updated":"2024-11-15 12:20:32.000000000","message":"Why is the iothread count parameter an integer when the number of iothreads is no longer relevant. If I\u0027m not mistaken, the code always uses iothread with index 1\n\nIf the VM has multiple disks (volumes for example) it makes sense to be able to split these disks between multiple iothreads.\n\nSome time ago we needed this feature in our version of nova-compute and we implemented it by allocating disks to iothreads by round robin mechanism","commit_id":"b20e730cca63e388cf661585d4c1c18f86f8f99c"},{"author":{"_account_id":34112,"name":"Kaloyan Kotlarski","email":"kaloyan.kotlarski@storpool.com","username":"kkotlarski"},"change_message_id":"1f208aec7fae863373ee43a2310c59afb6af7f5e","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"291bc9df_ad3afc10","updated":"2025-01-02 14:30:29.000000000","message":"recheck","commit_id":"b20e730cca63e388cf661585d4c1c18f86f8f99c"},{"author":{"_account_id":16207,"name":"ribaudr","display_name":"uggla","email":"rene.ribaud@gmail.com","username":"uggla","status":"Red Hat"},"change_message_id":"ea05e68a8374794aba9146dfa08102ebeeff8204","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":4,"id":"fd5c2914_7b5cc276","updated":"2025-07-04 13:11:16.000000000","message":"Thanks a lot for submitting this patch! I really appreciate your interest in contributing to Nova.\n\nApologies if I’m being a bit procedural here — I don’t want to discourage participation in any way — but I wanted to give you a heads-up that patches like this, especially when they touch non-trivial features, usually require going through the spec process before we can consider merging them.\n\nIn particular, features like the one you’re proposing are best discussed first during the Project Team Gathering (PTG), where core maintainers and contributors can align on the direction. After that, we typically expect a specification to be submitted and approved before working on the actual implementation. This helps everyone get on the same page and improves the chance of a successful merge.\n\nThere are already two related specs under discussion that may be of interest:\n\n- https://review.opendev.org/c/openstack/nova-specs/+/953940 (iothreads)\nand\n- https://review.opendev.org/c/openstack/nova-specs/+/953940 (multiqueue)\n\nA couple of suggestions:\n\n    I strongly encourage you to participate in the review of those specs, especially if your use case isn’t fully covered, your input would be very valuable.\n\n    We’d absolutely welcome your help with the implementation phase once the specs are approved! You might even be able to adapt your current patch to fit into that effort.\n\nThanks again for your contribution, it’s great to see new folks getting involved. Don’t hesitate to reach out if you have questions about the process or want guidance.","commit_id":"06cdbef356f3742047530ab418bb11b03b4582dc"},{"author":{"_account_id":16207,"name":"ribaudr","display_name":"uggla","email":"rene.ribaud@gmail.com","username":"uggla","status":"Red Hat"},"change_message_id":"5a68f15adc1d28704b8d6280f6adadab89e2f09e","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":4,"id":"b5cdd005_22f2f6b6","updated":"2025-07-04 13:57:22.000000000","message":"Thanks a lot for submitting this patch! I really appreciate your interest in contributing to Nova.\n\nApologies if I’m being a bit procedural here — I don’t want to discourage participation in any way — but I wanted to give you a heads-up that patches like this, especially when they touch non-trivial features, usually require going through the spec process before we can consider merging them.\n\nIn particular, features like the one you’re proposing are best discussed first during the Project Team Gathering (PTG), where core maintainers and contributors can align on the direction. After that, we typically expect a specification to be submitted and approved before working on the actual implementation. This helps everyone get on the same page and improves the chance of a successful merge.\n\nThere are already two related specs under discussion that may be of interest:\n\n-https://review.opendev.org/c/openstack/nova-specs/+/953940 (iothreads)\nand\n-https://review.opendev.org/c/openstack/nova-specs/+/953940 (multiqueue)\n\nA couple of suggestions:\n\nI strongly encourage you to participate in the review of those specs, especially if your use case isn’t fully covered, your input would be very valuable.\n\nWe’d absolutely welcome your help with the implementation phase once the specs are approved! You might even be able to adapt your current patch to fit into that effort.\n\nThanks again for your contribution, it’s great to see new folks getting involved. Don’t hesitate to reach out if you have questions about the process or want guidance.","commit_id":"06cdbef356f3742047530ab418bb11b03b4582dc"},{"author":{"_account_id":11604,"name":"sean mooney","email":"smooney@redhat.com","username":"sean-k-mooney"},"change_message_id":"34d61218284857ddee550f8693191209b03164da","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":4,"id":"201c144a_f565a6bb","updated":"2025-07-04 11:30:39.000000000","message":"just to note we have moved away form condiering config driven iothread entirly.\n\nthat was never our prefered approch but we are acvitly condiering flavor based iothread support https://review.opendev.org/c/openstack/nova-specs/+/953940/6/specs/2025.2/approved/static-vm-iothreads.rst","commit_id":"06cdbef356f3742047530ab418bb11b03b4582dc"},{"author":{"_account_id":11604,"name":"sean mooney","email":"smooney@redhat.com","username":"sean-k-mooney"},"change_message_id":"21c411550c0ef3c91579e50c790c68421178a6ba","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":4,"id":"61c5c378_dc2e8185","updated":"2025-07-04 11:28:50.000000000","message":"we hwave 2 spec desigin how iothread can be enabled in the future\n\nthis is not part of either of them and not something we can/shoudl do as a bugfix or specless bluepitnt","commit_id":"06cdbef356f3742047530ab418bb11b03b4582dc"},{"author":{"_account_id":35429,"name":"Biser Milanov","email":"biser.milanov@storpool.com","username":"sp-bmilanov"},"change_message_id":"ddcaebc9b170517980bc517d1d902facbb68dd51","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":4,"id":"534ee60d_f9bdba38","in_reply_to":"b5cdd005_22f2f6b6","updated":"2025-07-31 07:37:55.000000000","message":"Hi uggla,\n\nYes, I knew from the initial upload that we would have to describe and\ndiscuss the change before actually getting it merged. Not being merged\ncurrently is not an issue. It may even become obsolete if a version of\nhttps://review.opendev.org/c/openstack/nova-specs/+/953940 that works\nfor us is implemented.\n\nThanks for linking the other spec, I will take a look.","commit_id":"06cdbef356f3742047530ab418bb11b03b4582dc"}]}