)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":38368,"name":"Christian Ohanaja","display_name":"Christian Ohanaja","email":"cohanaja@nvidia.com","username":"cohanaja"},"change_message_id":"2eaa05a2ab13c3bb19a67e8adbce4ee0a9a60aed","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"971656c5_05d19d45","updated":"2025-09-24 17:59:23.000000000","message":"This adds a new stat to the recon logs for a sharder instance that keeps track of:\n1. The amount of time it took to run the process_broker function\n2. The total amount of time that the database has been sharding. If the sharder is the last one to cleave a DB (it moves the DB to SHARDED state), it\u0027ll be the final full runtime of the sharding process for that DB. (final sharder epoch - shard_range epoch) \n\nFor every db that the sharder instance went over, the recon logs will have an entry similar to\n\n\"\u003chash\u003e.db\":\n\n\"cycle_timings\": \n{\n    \"\u003chash\u003e.db\": [\n        0.6109530925750732,\n        44.17239737510681\n    ]\n}\n\nKeeping track of these metrics is part of the process of improving sharding times","commit_id":"4fbcfdaafe1051ffe2484a1a49ab4b510da450c0"}]}