igor-vodopyanov opened a new issue #7908: Reindexing that results in empty 
partitions doesn't work
URL: https://github.com/apache/incubator-druid/issues/7908
 
 
   ### Affected Version
   
   0.13.0-incubating-iap11
   
   ### Description
   
   Hi folks,
   I'm doing Hadoop reindexing with partitionSpec that looks like { "type": 
"hashed", "numShards": 100, "partitionDimensions": ["tenantId"]}. 
   If tenantId cardinality is high and each of 100 partitions obtains some rows 
then new segments are created, old are marked as unused, and everything goes 
fine. In a case when tenantId has not much distinct values and reindexing 
results in <100 partitions then I have an issue. New segments are created and 
are marked as used in druid_segments but old ones are still in place. And they 
are also marked as used druid_segments. If I delete old segments by doing 
DELETE /druid/coordinator/v1/datasources/{datasourceName}/segments/{segmentId} 
then I lose the data. If I delete new segments then it takes no effect other 
that marking segments as unused.
   The same happens if I use compaction task instead of Hadoop.
   
   I'm using Imply 2.8.19 that bundles Druid 0.13.0-incubating-iap11. Is this a 
known issue that was fixed in newer version?
   Are there other ways to partition and ideally have one partition per 
tentantId?

----------------------------------------------------------------
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.
 
For queries about this service, please contact Infrastructure at:
[email protected]


With regards,
Apache Git Services

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to