)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":27582,"name":"Simon Westphahl","email":"simon.westphahl@bmw.de","username":"simon.westphahl"},"change_message_id":"3dd5825968fc32daabc7e77fb99886148104b840","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":3,"id":"3bb8b6f7_d56bc6f6","updated":"2022-02-16 20:23:15.000000000","message":"recheck","commit_id":"bbc3ad1275aab1730f4e9bda3a5cd7cb4e872dbc"}],"zuul/scheduler.py":[{"author":{"_account_id":4146,"name":"Clark Boylan","email":"cboylan@sapwetik.org","username":"cboylan"},"change_message_id":"0ed8c34662f6f43b22b83eed811e666236333955","unresolved":true,"context_lines":[{"line_number":1886,"context_line":"                return"},{"line_number":1887,"context_line":"            if RECONFIG_LOCK_ID in tenant_lock.contenders():"},{"line_number":1888,"context_line":"                self.log.debug("},{"line_number":1889,"context_line":"                    \"Releasing tenant %s read lock due to pending reconfig\","},{"line_number":1890,"context_line":"                    tenant.name)"},{"line_number":1891,"context_line":"                self.wake_event.set()"},{"line_number":1892,"context_line":"                return"}],"source_content_type":"text/x-python","patch_set":4,"id":"8e15de58_0cf1d622","line":1889,"updated":"2022-02-21 17:23:03.000000000","message":"We aren\u0027t just releasing the lock we are also short circuiting pipeline processing for the tenant. Would it be better to log this as:\n\n  self.log.debug(\"Stopping tenant %s pipeline processing in order to release the tenant read lock.\", tenant.name)\n\nSeparately checking the contenders will involve ZK traffic. I expect that compared to long waits for the lock this is preferable but calling that out in case it represents a problem.","commit_id":"5516185a8ae6a7ff72250d0a98bc25e531e16eea"},{"author":{"_account_id":27582,"name":"Simon Westphahl","email":"simon.westphahl@bmw.de","username":"simon.westphahl"},"change_message_id":"4cae93d605c0c8541cc0e3b4358f56666a11eb8f","unresolved":true,"context_lines":[{"line_number":1886,"context_line":"                return"},{"line_number":1887,"context_line":"            if RECONFIG_LOCK_ID in tenant_lock.contenders():"},{"line_number":1888,"context_line":"                self.log.debug("},{"line_number":1889,"context_line":"                    \"Releasing tenant %s read lock due to pending reconfig\","},{"line_number":1890,"context_line":"                    tenant.name)"},{"line_number":1891,"context_line":"                self.wake_event.set()"},{"line_number":1892,"context_line":"                return"}],"source_content_type":"text/x-python","patch_set":4,"id":"bef727c2_e041f1c7","line":1889,"in_reply_to":"8e15de58_0cf1d622","updated":"2022-02-22 06:24:37.000000000","message":"Yes, that might be better. I think I\u0027ll also leave the read lock stuff out of the message as that\u0027s just an implementation detail.\n\nRe. ZK traffic: yep, this needs a few ZK requests, but as there shouldn\u0027t be too many contenders at any given time I think it won\u0027t have a significant performance impact.","commit_id":"5516185a8ae6a7ff72250d0a98bc25e531e16eea"}]}