lokesh-lingarajan commented on a change in pull request #12321:
URL: https://github.com/apache/druid/pull/12321#discussion_r828545572
##########
File path: web-console/src/views/ingestion-view/ingestion-view.tsx
##########
@@ -244,6 +247,8 @@ ORDER BY "rank" DESC, "created_time" DESC`;
hiddenSupervisorColumns: new LocalStorageBackedVisibility(
LocalStorageKeys.SUPERVISOR_TABLE_COLUMN_SELECTION,
),
+
+ preferOverlord: true,
Review comment:
@gianm , you are right. I think the confusion was between the APIs
"/druid/indexer/v1/task" and "/druid/indexer/v1/tasks". First one for posting
new tasks(thru ingestion specs) and the second one that is being used for
fetching the task list. Second one is going to the backend metadata store.
Please ignore this PR.
But the underlying problem is that the data model for druid_tasks table has
task status embedded inside "status_payload" and we need to fetch all the tasks
to UI in order to filter them as running, success and failed. At our current
scale we are fetching a lot of data from RDS and this is causing network
throughput increase and affecting performance of the system. Ideal way would be
to remodel the table to have status as a field which can queried and filter at
metadata store itself as explained
https://github.com/apache/druid/issues/12318.
Hope this helps.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
