)]}'
{"roles/oooci-build-images/tasks/main.yaml":[{"author":{"_account_id":3153,"name":"Emilien Macchi","email":"emilien@redhat.com","username":"emilienm"},"change_message_id":"da6ab75da637794cca5007feba3394f5ea335c7b","unresolved":false,"context_lines":[{"line_number":25,"context_line":"  register: result"},{"line_number":26,"context_line":"  until: result is success"},{"line_number":27,"context_line":"  retries: 3"},{"line_number":28,"context_line":"  delay: 60"},{"line_number":29,"context_line":""},{"line_number":30,"context_line":"# The python_v fact is used with vars/common.yaml to set the right"},{"line_number":31,"context_line":"# overcloud-images file in the build-images.sh.j2 template"}],"source_content_type":"text/x-yaml","patch_set":3,"id":"bf51134e_227d246e","line":28,"updated":"2020-07-23 21:19:27.000000000","message":"a delay of 1 min seems really long. Why would you wait so much between the retries? it seems to be a waste of time. I would rather use a smaller value, like 5 seconds maybe.","commit_id":"201f4398c61f0f6ff614f2fe42af6e4a363b00ec"},{"author":{"_account_id":3153,"name":"Emilien Macchi","email":"emilien@redhat.com","username":"emilienm"},"change_message_id":"fc85a9251160057ea8d132051984ff4c72b608e1","unresolved":false,"context_lines":[{"line_number":25,"context_line":"  register: result"},{"line_number":26,"context_line":"  until: result is success"},{"line_number":27,"context_line":"  retries: 3"},{"line_number":28,"context_line":"  delay: 60"},{"line_number":29,"context_line":""},{"line_number":30,"context_line":"# The python_v fact is used with vars/common.yaml to set the right"},{"line_number":31,"context_line":"# overcloud-images file in the build-images.sh.j2 template"}],"source_content_type":"text/x-yaml","patch_set":3,"id":"9f560f44_fe22d000","line":28,"in_reply_to":"9f560f44_0b9e6853","updated":"2020-07-29 18:12:52.000000000","message":"I would do more retries and less delay between them.\n\nSo let\u0027s say we use a delay of 10 seconds, and let a downtime up to 5 min (which is IMO high enough for CI); it means 600/10\u003d60\n\nretries: 60\ndelay: 10","commit_id":"201f4398c61f0f6ff614f2fe42af6e4a363b00ec"},{"author":{"_account_id":8449,"name":"Marios Andreou","email":"marios.andreou@gmail.com","username":"marios"},"change_message_id":"13341fbe97b8de9b4bb092d202dc5e99ac001f40","unresolved":false,"context_lines":[{"line_number":25,"context_line":"  register: result"},{"line_number":26,"context_line":"  until: result is success"},{"line_number":27,"context_line":"  retries: 3"},{"line_number":28,"context_line":"  delay: 60"},{"line_number":29,"context_line":""},{"line_number":30,"context_line":"# The python_v fact is used with vars/common.yaml to set the right"},{"line_number":31,"context_line":"# overcloud-images file in the build-images.sh.j2 template"}],"source_content_type":"text/x-yaml","patch_set":3,"id":"bf51134e_788ce8d6","line":28,"in_reply_to":"bf51134e_227d246e","updated":"2020-07-24 07:44:46.000000000","message":"+1 and if you have a smaller delay you might also consider adding another retry (though 3 is probably enough)","commit_id":"201f4398c61f0f6ff614f2fe42af6e4a363b00ec"},{"author":{"_account_id":24162,"name":"Sorin Sbârnea","display_name":"zbr","email":"ssbarnea@redhat.com","username":"ssbarnea","status":"do not feed the troll 🥕"},"change_message_id":"a1a34e19ca65a1c80b61e045d2ed70add5392ebc","unresolved":false,"context_lines":[{"line_number":25,"context_line":"  register: result"},{"line_number":26,"context_line":"  until: result is success"},{"line_number":27,"context_line":"  retries: 3"},{"line_number":28,"context_line":"  delay: 60"},{"line_number":29,"context_line":""},{"line_number":30,"context_line":"# The python_v fact is used with vars/common.yaml to set the right"},{"line_number":31,"context_line":"# overcloud-images file in the build-images.sh.j2 template"}],"source_content_type":"text/x-yaml","patch_set":3,"id":"9f560f44_0b9e6853","line":28,"in_reply_to":"bf51134e_227d246e","updated":"2020-07-29 16:58:52.000000000","message":"We are talking about a \u003e1GB image, and the delay should be proportional to the expected time related for a successful download. \n\nIf we put 5s, it means this job will fail if networking has a glitch of 15s, which is clearly not heard of.\n\nThe goal here is to be able to survive networking glitches, like rdo server restarts!\n\nI would argue that even the current 3x60\u003d180s may be too low. \n\nI would be more confident with a set of values that would allow to survive even 10min outages.\n\nAnyway, you are welcomed to edit the patch and put whatever you want. It is already too late in the day for me and all I know for sure is that every day without any retries means more opportunities for failed builds.","commit_id":"201f4398c61f0f6ff614f2fe42af6e4a363b00ec"}]}