)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":36080,"name":"Erkin Mussurmankulov","display_name":"Eric","email":"mangust404@gmail.com","username":"mongoose404","status":"PS Cloud services employee"},"change_message_id":"3de3ee0875b73a7cb501a2afa575c13d101c7dc3","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"75afc8d3_aa831b2e","updated":"2025-12-23 16:10:47.000000000","message":"Hello, Wu!\nThis solution looks brilliant!\n\nInside the database container I can see that /dev/shm is the same size as the host\u0027s /dev/shm, so `ipc_mode\u003dhost` works as expected.\n\nI was trying to replicate the bug with overwhelming /dev/shm with the default 64M on my dev server, but with no luck. It seems that it can be replicated only on production database with big dataset and a lot of workers. So unfortunately I can\u0027t test this solution on real instances.\n\nBut I think it WILL work.","commit_id":"2312a8bfeba8eb3f84441d8090d8a285cb23cd15"},{"author":{"_account_id":31737,"name":"Hirotaka Wakabayashi","email":"hiwkby@yahoo.com","username":"hiwkby"},"change_message_id":"83f0a135c2c5d7d7963f5f56da84a4f0499ba5c4","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"f0a53954_05747742","updated":"2025-12-29 23:53:22.000000000","message":"Hello Wu, Thanks for the PR! LGTM.","commit_id":"124d89a896d80e6ae9d6cc9b3792de9cc8f730ee"}]}
