Retro Eye care Haitian Deep Dark Default

The database which is the value of the event tracing property event_trace_rdb_url will automatically create two tables JOB_EXECUTION_LOG and JOB_STATUS_TRACE_LOG and several indexes.

JOB_EXECUTION_LOG Columns

Column name Column type Required Describe
id VARCHAR(40) Yes Primary key
job_name VARCHAR(100) Yes Job name
task_id VARCHAR(1000) Yes Task name, create new tasks every time the job runs.
hostname VARCHAR(255) Yes Hostname
ip VARCHAR(50) Yes IP
sharding_item INT Yes Sharding item
execution_source VARCHAR(20) Yes Source of job execution. The value options are NORMAL_TRIGGER, MISFIRE, FAILOVER.
failure_cause VARCHAR(2000) No The reason for execution failure
is_success BIT Yes Execute successfully or not
start_time TIMESTAMP Yes Job start time
complete_time TIMESTAMP No Job end time

JOB_EXECUTION_LOG records the execution history of each job. There are two steps:

  1. When the job is executed, program will create one record in the JOB_EXECUTION_LOG, and all fields except failure_cause and complete_time are not empty.
  2. When the job completes execution, program will update the record, update the columns of is_success, complete_time and failure_cause(if the job execution fails).

JOB_STATUS_TRACE_LOG Columns

Column name Column type Required Describe
id VARCHAR(40) Yes Primary key
job_name VARCHAR(100) Yes Job name
original_task_id VARCHAR(1000) Yes Original task name
task_id VARCHAR(1000) Yes Task name
slave_id VARCHAR(1000) Yes Server’s name of executing the job. The valve is server’s IP.
execution_type VARCHAR(20) Yes Type of job execution, the value options are NORMAL_TRIGGER, MISFIRE, FAILOVER.
sharding_item VARCHAR(255) Yes Collection of sharding item, multiple sharding items are separated by commas.
state VARCHAR(20) Yes State of job execution, the value options are TASK_STAGING, TASK_RUNNING, TASK_FINISHED, TASK_KILLED, TASK_LOST, TASK_FAILED, TASK_ERROR.
message VARCHAR(2000) Yes Message
creation_time TIMESTAMP Yes Create time

JOB_STATUS_TRACE_LOG record the job status changes. Through the task_id of each job, user can query the life cycle and running track of the job status change.