[
https://issues.apache.org/jira/browse/KYLIN-5852?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
pengfei.zhan updated KYLIN-5852:
--------------------------------
Description: There are two steps need to handle high priority jobs: 1)
write the ready task into the job_lock table on the dataloading master node; 2)
read the task from the job_lock table to execute it on the dataloading slave
node. The priority of tasks is correctly managed by the master node; however,
this property is not serialized into the job_lock table, so the slave node
cannot order the tasks correctly. (was: Two steps need handle high priority
jobs: 1) write the ready task into the job_lock table in the dataloading master
node; 2) read the task from job_lock table to execute in the dataloading slave
node. The priority of tasks is correctly handled by the master, however, the
priority property does not serialized into the job_lock, then the slave can not
order the tasks in the correct way.)
> Support priority for job scheduler
> ----------------------------------
>
> Key: KYLIN-5852
> URL: https://issues.apache.org/jira/browse/KYLIN-5852
> Project: Kylin
> Issue Type: Improvement
> Components: Job Engine
> Affects Versions: 5.0.0
> Reporter: pengfei.zhan
> Assignee: pengfei.zhan
> Priority: Major
> Fix For: 5.0.0
>
>
> There are two steps need to handle high priority jobs: 1) write the ready
> task into the job_lock table on the dataloading master node; 2) read the task
> from the job_lock table to execute it on the dataloading slave node. The
> priority of tasks is correctly managed by the master node; however, this
> property is not serialized into the job_lock table, so the slave node cannot
> order the tasks correctly.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
