)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":13252,"name":"Dr. Jens Harbott","display_name":"Jens Harbott (frickler)","email":"frickler@offenerstapel.de","username":"jrosenboom"},"change_message_id":"b737148f98d01b3e2dcd4a091831b2f0f0228640","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":8,"id":"193f4ed2_3f809248","updated":"2026-08-17 14:48:51.000000000","message":"I have mixed feelings about this, I can understand that infra folx want us to optimize our CI jobs, otoh there is an argument to be made for jobs to follow that same path that actual deployers would take. the bigger the delta the more new issues we may run into\n\nalso afaict the horizon plugins are actually broken, we\u0027be been dropping them downstream for a while","commit_id":"25c2df8dc2ccd7b0be93417d536a96d715362a44"},{"author":{"_account_id":22629,"name":"Michal Nasiadka","email":"mnasiadka@gmail.com","username":"mnasiadka"},"change_message_id":"1357c38b1c1c322ef4bc77d207d4f4aafaeade05","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":8,"id":"011f2629_7e114b78","in_reply_to":"14a4dcae_a15b5aae","updated":"2026-08-18 10:07:44.000000000","message":"Done","commit_id":"25c2df8dc2ccd7b0be93417d536a96d715362a44"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"4c95e793a9a29693b41a0e55180fc369bc4f673d","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":8,"id":"50a28dee_fb5e9dd8","in_reply_to":"193f4ed2_3f809248","updated":"2026-08-17 16:54:05.000000000","message":"Agree, let\u0027s just point it to github instead tarballs and that\u0027s it.","commit_id":"25c2df8dc2ccd7b0be93417d536a96d715362a44"},{"author":{"_account_id":22629,"name":"Michal Nasiadka","email":"mnasiadka@gmail.com","username":"mnasiadka"},"change_message_id":"9a0e6ef0fc7de2c0bf9e87d893a4d14b48e8030d","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":8,"id":"14a4dcae_a15b5aae","in_reply_to":"4ec7415f_1978479e","updated":"2026-08-18 04:59:51.000000000","message":"Zuul actively keeps that cache fresh and it costs less cpu-wise to use that instead of fetching from opendev.org (Gitea servers) - shallow clones could be better, but I don\u0027t think we\u0027re doing that at the moment.","commit_id":"25c2df8dc2ccd7b0be93417d536a96d715362a44"},{"author":{"_account_id":22629,"name":"Michal Nasiadka","email":"mnasiadka@gmail.com","username":"mnasiadka"},"change_message_id":"4974dc98ee51a59c1493c57f9f02d6d8c5e3afc0","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":8,"id":"e72f8e48_0206d826","in_reply_to":"50a28dee_fb5e9dd8","updated":"2026-08-17 17:11:11.000000000","message":"I agree to disagree - we probably should point to GitHub, but when GitHub is down - we\u0027ll have the same problem, tarballs is a single server backed by AFS, and the tarballs upload might fail just like it failed today - so I think at least our publish jobs should use Zuul cached repositories (which puts less strain on tarballs and Gitea) - but we should have some jobs that validate if everything works from a user perspective. I think we should discuss that on this weeks meeting.","commit_id":"25c2df8dc2ccd7b0be93417d536a96d715362a44"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"3fdc18d46b002c38d274f05aa90d8d80d1e72cdf","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":8,"id":"4ec7415f_1978479e","in_reply_to":"e72f8e48_0206d826","updated":"2026-08-17 17:17:13.000000000","message":"Well, i don\u0027t remember if github was down. And, there are opendev repos also ? What about old cache ?","commit_id":"25c2df8dc2ccd7b0be93417d536a96d715362a44"}]}