Successful ingest, no data in query

[Filip Dziuba]

Hi,
I am trying to reindex data and change query granularity. I am reading from Druid data source A and writing to data source B.
Ingestion is separated into 20 subtasks with 10 running at the same time. One subtask usually fails but on retry it succeeded so the whole index_parallel task is successful. 

I can see that segment and partitions are created, Druid UI see the number of rows, but query return no data. There are no errors on any node.

I am quite stuck, the spec work on smaller data set.
The only worrying thing I can see is that when one subtask fails the partition it was supposed to create is never created. The retry will create new partition. So I can have set of 20 partitions 0-20 with 12th partition missing, so count is OK, but they are not in sequence.

Is this a bug or should druid handle partitions with missing numbers?

[Vaibhav Vaibhav]

Hi Filip ,

The non-contiguous segment partitionID seems to be the case of this issue. Its a BUG and there is a pull request for this -

https://github.com/apache/druid/pull/11189

I think the fix is merged in the druid-0.22 release.

Regards,
Vaibhav

--
You received this message because you are subscribed to the Google Groups "Druid User" group.
To unsubscribe from this group and stop receiving emails from it, send an email to druid-user+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/druid-user/9632c5b8-78dc-4690-9b39-6ff60f4d86e9n%40googlegroups.com.

[Filip Dziuba]

Thanks for help, I will see if this is will fix it.