[ https://issues.apache.org/jira/browse/AIRFLOW-3300?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16676990#comment-16676990 ]
Ash Berlin-Taylor commented on AIRFLOW-3300: -------------------------------------------- Might be fixed by AIRFLOW-2703, but a deadlock is possibly a sign of a bigger issue. > Frequent crash of scheduler while interacting with Airflow Metadata (Mysql) > --------------------------------------------------------------------------- > > Key: AIRFLOW-3300 > URL: https://issues.apache.org/jira/browse/AIRFLOW-3300 > Project: Apache Airflow > Issue Type: Bug > Reporter: Tanuj Gupta > Priority: Major > > It's been very frequent when scheduler tries to update the task instance > table and ends up with scheduler crash due to deadlock occurrence. Following > is the stack-trace for the same - > > {noformat} > sqlalchemy.exc.OperationalError: (_mysql_exceptions.OperationalError) (1213, > 'Deadlock found when trying to get lock; try restarting transaction') [SQL: > u'UPDATE task_instance, dag_run SET task_instance.state=%s WHERE > task_instance.dag_id IN (%s, %s, %s, %s, %s) AND task_instance.state IN (%s, > %s) AND dag_run.dag_id = task_instance.dag_id AND dag_run.execution_date = > task_instance.execution_date AND dag_run.state != %s'] [parameters: (None, > 'org_test0802_h9zrva', > '27bd514b5ab9854b0a494110_45aa7868_1799_4046_ad20_f35e3de1a4ec_p8bkfg', > 'org_e2e_trainman1528457521430', 'org_blockerretryissue_6kiafp', > 'org_e2e_compute_v21540610294106_svnfmr', u'queued', u'scheduled', > u'running')] (Background on this error at: > http://sqlalche.me/e/e3q8){noformat} > -- This message was sent by Atlassian JIRA (v7.6.3#76005)