)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":1004,"name":"Mohammed Naser","email":"mnaser@vexxhost.com","username":"mnaser"},"change_message_id":"fd0e24b52618e403d302d1b9eda8b701132f6d83","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"b057ee7e_eb70a8b6","updated":"2024-06-25 12:28:39.000000000","message":"-1 from me but can you talk more about the vswitchd crash loop issues?  we seen those sometimes here and we\u0027d like to fix them.","commit_id":"a84e58add4a672894ca6d9c77e94cbb853a96c49"},{"author":{"_account_id":1004,"name":"Mohammed Naser","email":"mnaser@vexxhost.com","username":"mnaser"},"change_message_id":"b1a4672efa8aed7c33c7c079124d505228fa87cf","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"d4ebfeef_7cab24dc","updated":"2024-06-26 17:24:06.000000000","message":"Hi Vlad, that is super useful, the init parent process change was that somewhere in the chart or?","commit_id":"a84e58add4a672894ca6d9c77e94cbb853a96c49"},{"author":{"_account_id":3009,"name":"Vladimir Kozhukalov","email":"kozhukalov@gmail.com","username":"kozhukalov"},"change_message_id":"54503b97daa054958c558a48999539ca0530125b","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"a8063a51_4ef760b9","updated":"2024-05-30 10:31:57.000000000","message":"Looks like the original commit you are referring to https://review.opendev.org/c/openstack/openstack-helm-infra/+/866478 was merged with the strong arguments that vswitchd always needs ovs-db running on the same node and they must be of the same version. So, when daemon set restarts pods during upgrade we are sure that if the pod starts they everything is consistent. \n\nEven taking into account this containerd \u003e1.7 issue and pod crash loop, frankly, I don\u0027t see any reason why we should again have two daemonsets instead of one.","commit_id":"a84e58add4a672894ca6d9c77e94cbb853a96c49"},{"author":{"_account_id":3009,"name":"Vladimir Kozhukalov","email":"kozhukalov@gmail.com","username":"kozhukalov"},"change_message_id":"90cca81264c483e0b65e1c9fbb4a016975d93b1d","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"a6cf2786_df8c6e5a","in_reply_to":"b057ee7e_eb70a8b6","updated":"2024-06-26 17:22:55.000000000","message":"I was talking about this recent change in containerd https://github.com/containerd/containerd/pull/6866/commits/46773c341796748a7057e4f03826e738bc6778ec#diff-73ee35f1fd7c9ae61938ea0ac335514aa634e62aa88696ee0be8e2c144fb5155R144 which seems might have affected vswitchd containers. It was something like shim was cleaned up for some reason but not the container filesystem where the previous pid file prevented vswitchd to start it again. I wasn\u0027t directly involved into the debugging this. It seems that using init parent process helped to resolve the issue.","commit_id":"a84e58add4a672894ca6d9c77e94cbb853a96c49"},{"author":{"_account_id":3009,"name":"Vladimir Kozhukalov","email":"kozhukalov@gmail.com","username":"kozhukalov"},"change_message_id":"783293c0e4a0a4f481a74a92e6233de04392f059","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"a3a12384_be15973c","in_reply_to":"d4ebfeef_7cab24dc","updated":"2024-06-26 18:23:31.000000000","message":"Yes. We merged these two changes regarding this issue\n\nhttps://review.opendev.org/c/openstack/openstack-helm-images/+/920324\nhttps://review.opendev.org/c/openstack/openstack-helm-infra/+/920768","commit_id":"a84e58add4a672894ca6d9c77e94cbb853a96c49"}]}