)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"afbd759ef8a815f4ae6cb19cd0ccf5075328318f","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"2c8e149d_07257cb4","updated":"2025-12-22 17:42:55.000000000","message":"@cardoe@cardoe.com I\u0027d not mind someone adding ironic-python-agent-minimal with literally only core dependencies of IPA (maybe even without things like ipmitool) and then maybe also ironic-python-agent-full for cases like this one.","commit_id":"1579279deb0a101bd8357960ce46c651d627d35d"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"c0d6f9461434c26ee96ef9a529302f7f5653decf","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"3185c1d5_ab73191c","updated":"2025-12-10 14:03:31.000000000","message":"And there seem to be an element for that: https://opendev.org/openstack/ironic-python-agent-builder/src/branch/master/dib/element/ironic-python-agent-podman","commit_id":"1579279deb0a101bd8357960ce46c651d627d35d"},{"author":{"_account_id":5890,"name":"Doug Goldstein","email":"cardoe@cardoe.com","username":"cardoe"},"change_message_id":"b916e68c166615c36680fded201a175ad53e091f","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":1,"id":"032447d9_b700e00a","updated":"2025-12-20 16:52:06.000000000","message":"For me personally I\u0027d like to see a small dib be the ironic-python-agent-ramdisk and then we have a handful of other dibs to add things. Then we can have a dib that\u0027s just inclusive of the kitchen sink.\n\nFor example today I don\u0027t even use ironic-python-agent-ramdisk so that I can turn off the dhcp-all-interfaces because it literally adds 4 minutes to the IPA boot time... every.... single... time...","commit_id":"1579279deb0a101bd8357960ce46c651d627d35d"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"585c863822d4ef7fe0e9d5d5714c23f6447e2327","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"b64fd3e8_b1c96e77","updated":"2025-12-10 12:03:33.000000000","message":"I\u0027m very hesitant to add these by default, unless you can proof that the size increase is negligible.","commit_id":"1579279deb0a101bd8357960ce46c651d627d35d"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"81b04e57a12b28d6efc38c17f967d6f1826b2d98","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":1,"id":"4d42763b_6bdc11aa","in_reply_to":"032447d9_b700e00a","updated":"2026-07-15 13:50:52.000000000","message":"... Only 4 minutes?\n\nSeriously though, I do like the idea of splitting flavors apart so there is a minimal case, and then a more expansive case. If operators only want to use minimal or maximal image, that is okay. They know what they are doing in their environment, just like you\u0027ve had to route around issues such as..  your 4 minute boots.\n\nThat is not to say there is not room for improvement, but its a \"is it worth the squeeze\" in that case. For example, I\u0027ve seen some cases over the years where folk shave had 20+ nics on a single host mostly all cabled, and that created absurd boot delays. In the end, we did the minimal. There could be room to improve that overall, but again, its use case/model dependent, and that is okay.","commit_id":"1579279deb0a101bd8357960ce46c651d627d35d"},{"author":{"_account_id":4571,"name":"Steve Baker","email":"sbaker@redhat.com","username":"steve-stevebaker"},"change_message_id":"3df3d5d2c2897bef993cb74578986bab83e388b9","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"cce61410_b3fef19c","in_reply_to":"3185c1d5_ab73191c","updated":"2025-12-11 00:51:40.000000000","message":"The element is a bit different, it allows container based steps to be defined which also has podman as a dependency. Since podman is invoked by standby.py I\u0027m hoping we can treat it as a core dependency, if the increase in size is acceptable.","commit_id":"1579279deb0a101bd8357960ce46c651d627d35d"},{"author":{"_account_id":5890,"name":"Doug Goldstein","email":"cardoe@cardoe.com","username":"cardoe"},"change_message_id":"eca7ec509dcbb1e1b472d43ef38fc331ca9e5b87","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":1,"id":"7e2f2eb4_60ccaba0","in_reply_to":"4d42763b_6bdc11aa","updated":"2026-07-16 17:34:21.000000000","message":"Just give me an actual minimal and let me build up from there. Let\u0027s have a \"default\" and it can be what we really recommend people use. But I\u0027m looking at multiple operators and they don\u0027t use ironic-python-agent-ramdisk with various comments as to why.","commit_id":"1579279deb0a101bd8357960ce46c651d627d35d"},{"author":{"_account_id":4571,"name":"Steve Baker","email":"sbaker@redhat.com","username":"steve-stevebaker"},"change_message_id":"f015e4f6c2d5db6333aeaaeddcc65536de1d8663","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":1,"id":"8580960c_0f3c6d22","in_reply_to":"7e2f2eb4_60ccaba0","updated":"2026-07-16 22:50:51.000000000","message":"OK, check out the previous change in this series: https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/997659","commit_id":"1579279deb0a101bd8357960ce46c651d627d35d"},{"author":{"_account_id":4571,"name":"Steve Baker","email":"sbaker@redhat.com","username":"steve-stevebaker"},"change_message_id":"5253129726d1f2dde3d6b4a9e0a40ea8f2048859","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"15c6cf61_0f1ee7e6","in_reply_to":"8580960c_0f3c6d22","updated":"2026-07-29 22:15:09.000000000","message":"Done","commit_id":"1579279deb0a101bd8357960ce46c651d627d35d"},{"author":{"_account_id":4571,"name":"Steve Baker","email":"sbaker@redhat.com","username":"steve-stevebaker"},"change_message_id":"3df3d5d2c2897bef993cb74578986bab83e388b9","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"787268ea_1742f1e6","in_reply_to":"b64fd3e8_b1c96e77","updated":"2025-12-11 00:51:40.000000000","message":"The size increase for centos9 is 29M: 495M [1] vs 466M[2]\n\n[1] https://zuul.opendev.org/t/openstack/build/55e71f54806e431b91da42ed916b13a9/log/job-output.txt#5387\n[2] https://zuul.opendev.org/t/openstack/build/c6661f0cbfe54f09b68de552ca10da2c/log/job-output.txt#5209\n\nMy goal is to demonstrate a bootc install using as much upstream metal3 artefacts as practical.","commit_id":"1579279deb0a101bd8357960ce46c651d627d35d"},{"author":{"_account_id":4571,"name":"Steve Baker","email":"sbaker@redhat.com","username":"steve-stevebaker"},"change_message_id":"1ab3290921efa3daea4bf1948100962ec4357591","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"71f4992d_473ed1a0","updated":"2026-07-14 01:09:16.000000000","message":"I\u0027m proposing this again at the end of an image size reduction series which shrinks the image significantly more than this adds.","commit_id":"8b1e29c5d50ba9d8ff8546ad0bd889c0f509ae52"},{"author":{"_account_id":4571,"name":"Steve Baker","email":"sbaker@redhat.com","username":"steve-stevebaker"},"change_message_id":"9bb32ea16741f7229d3471856c8b3391cb267c12","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":3,"id":"718b6e0a_2ada5bff","updated":"2026-07-16 03:57:53.000000000","message":"I have un-abandoned this change and put it at the end of a series which reduces the image size by 177M (35%!).\n\nThis change adds ~29M and triggered a request for a minimal image. On the basis of the size savings could we please decouple that discussion and revisit this change with fresh eyes? @dtantsur@protonmail.com @cardoe@cardoe.com","commit_id":"82487843d1850be93687f1a27e0c0de9f3edc648"},{"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":"8ed52c85964029c48d9061144246a7e1341832f5","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":5,"id":"d93d8fa8_8afff813","updated":"2026-07-29 22:34:30.000000000","message":"I agree with the earlier comments this shouldn\u0027t be included in IPA ramdisks by default.\n\nFolks not using the ContainerHardwareManager or bootc should not be subjected to the security risks they contain that only exist when podman or docker is installed.\n\nI understand metal3 wants to run CI with OpenStack-generated artifacts, but I don\u0027t think this is worth the tradeoff for operators that consume these. \n\nPlease do not merge this change unless we work through consensus in the larger team.","commit_id":"b8b145b685f77aa3698f661c021fc6eb384d3fc9"},{"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":"e2b9e2f7ba27946c259daf99d1afef9d616e438b","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":5,"id":"8efe63c4_0ec61a9a","updated":"2026-08-13 21:24:02.000000000","message":"With the publication of OSSN-0107, my concern about this is lessened. I still think we\u0027re better off with a minimal ramdisk, but I don\u0027t hold the opinion strongly enough to get in your way.","commit_id":"b8b145b685f77aa3698f661c021fc6eb384d3fc9"},{"author":{"_account_id":4571,"name":"Steve Baker","email":"sbaker@redhat.com","username":"steve-stevebaker"},"change_message_id":"e46f64e584cd0c3abf8c909e0160d0768b0f1f64","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":5,"id":"fad55486_17973c5a","in_reply_to":"d93d8fa8_8afff813","updated":"2026-07-30 00:03:36.000000000","message":"Would it help if we started publishing ironic-python-agent-base images[1] and signalling that this would be considered a default minimal image?\n\n[1] https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/997659","commit_id":"b8b145b685f77aa3698f661c021fc6eb384d3fc9"},{"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":"d57fd4c2f3179d46aacc99eb8d81a2e426d6a424","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":5,"id":"3484639a_d0a09614","in_reply_to":"fad55486_17973c5a","updated":"2026-08-06 16:22:07.000000000","message":"I don\u0027t really think so. I\u0027d prefer any operators using advanced features that require podman/docker build their own ramdisks.\n\nThere is a danger in making advanced Ironic features easily accessible, especially ones that have security implications.\n\nI\u0027m not sure the best way to resolve this disagreement, maybe a straight vote in a meeting or something?","commit_id":"b8b145b685f77aa3698f661c021fc6eb384d3fc9"}]}