)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":11166,"name":"Lukas Bezdicka","email":"social@v3.sk","username":"xbezdick"},"change_message_id":"fdd0337a9510bc6992b658db4370fe35f89f7793","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"72891507_a0922542","updated":"2021-11-10 14:06:42.000000000","message":"We want to restart docker and just ensure that only containers that were running before restart are running after. The restart of docker triggers all the containers to start which we don\u0027t want.\n\nWe don\u0027t want to stop docker.","commit_id":"6bf41fecb346d66e359bda3aeea84bd51f11cb1e"},{"author":{"_account_id":26343,"name":"Jose Luis Franco","email":"jfrancoa@redhat.com","username":"jfrancoa"},"change_message_id":"0506f9854d4be6f862cc55c02ccbd6d618529adf","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":4,"id":"03e587fe_66676b21","updated":"2021-11-16 06:42:35.000000000","message":"Is this still required after finding out what was the culprit for https://bugzilla.redhat.com/show_bug.cgi?id\u003d2022519 ?","commit_id":"1564f792f09c114b32c9c6521802cc73994027ef"}],"deployment/iscsid/iscsid-container-puppet.yaml":[{"author":{"_account_id":11090,"name":"Sergii Golovatiuk","email":"sgolovat@redhat.com","username":"holser"},"change_message_id":"721262f4c9c831f8e62a937a12e5b14f2d466a7d","unresolved":true,"context_lines":[{"line_number":248,"context_line":"                    # Get list of running containers"},{"line_number":249,"context_line":"                    RUNNING\u003d\"$( docker ps --format \u0027{{ \u0027{{\u0027 }}.Names{{ \u0027}}\u0027 }}\u0027 )\""},{"line_number":250,"context_line":"                    # Restart docker"},{"line_number":251,"context_line":"                    systemctl restart docker"},{"line_number":252,"context_line":"                    # Compare running containers now vs before"},{"line_number":253,"context_line":"                    TO_STOP\u003d\"$(grep -v -f \u003c(echo \"${RUNNING}\")  \u003c(docker ps --format \u0027{{ \u0027{{\u0027 }}.Names{{ \u0027}}\u0027 }}\u0027))\""},{"line_number":254,"context_line":"                    # Check if we need to stop anything and stop it"}],"source_content_type":"text/x-yaml","patch_set":2,"id":"3e06e8e5_3babda34","side":"PARENT","line":251,"updated":"2021-11-10 14:27:34.000000000","message":"The problem is that containers don\u0027t start immediately. We need some time to start them or a loop that checks that there are no containers that are about to start. Once all of them are started we can compare and stop containers that were not run before. \n\nTL;DR sleep 5 fixes the issue","commit_id":"bfe00e70583357a864ce49b37338b18b64c82af7"},{"author":{"_account_id":26343,"name":"Jose Luis Franco","email":"jfrancoa@redhat.com","username":"jfrancoa"},"change_message_id":"33f0c3d9688050a8105e13d5e19bfa99931f48a2","unresolved":true,"context_lines":[{"line_number":249,"context_line":"                    RUNNING\u003d\"$( docker ps --format \u0027{{ \u0027{{\u0027 }}.Names{{ \u0027}}\u0027 }}\u0027 )\""},{"line_number":250,"context_line":"                    # Restart docker"},{"line_number":251,"context_line":"                    systemctl restart docker"},{"line_number":252,"context_line":"                    sleep 5"},{"line_number":253,"context_line":"                    # Compare running containers now vs before"},{"line_number":254,"context_line":"                    TO_STOP\u003d\"$(grep -v -f \u003c(echo \"${RUNNING}\")  \u003c(docker ps --format \u0027{{ \u0027{{\u0027 }}.Names{{ \u0027}}\u0027 }}\u0027))\""},{"line_number":255,"context_line":"                    # Check if we need to stop anything and stop it"}],"source_content_type":"text/x-yaml","patch_set":4,"id":"aecee21f_2dde4183","line":252,"updated":"2021-11-16 14:25:03.000000000","message":"If possible I would prefer if we implement a busy waiting instead of sleeping 5 seconds by default... if we leave it like this we will be adding 5 seconds for free to every node we run the hybrid state in, which translates in several seconds in big environments.\nIf we actively check if the service is running (or perform a ping to the service like done here: https://stackoverflow.com/a/69921248 ) and wait if it isn\u0027t will turn into most cases a 1 second wait (or directly continueing) and 5 seconds in the worst case.\nI know this approach is more tedious, but it will save us upgrade time.","commit_id":"1564f792f09c114b32c9c6521802cc73994027ef"}]}