leerho commented on issue #6743: IncrementalIndex generally overestimates theta 
sketch size
URL: 
https://github.com/apache/incubator-druid/issues/6743#issuecomment-486759807
 
 
   @himanshug @gianm 
   
   > I looked into this a long time ago and one way was hacking it was to use 
"MemoryRegion" and "MemoryRequest" as in 
https://github.com/himanshug/druid/blob/growable_aggregator_final/extensions/datasketches/src/main/java/io/druid/query/aggregation/datasketches/theta/SketchResizableBufferAggregator.java#L120
 (as you might guess this is based on pretty old version of DS library :) ) .
   
   Those mechanisms still exist in the current library, the MemoryRequest was 
just renamed 
[MemoryRequestServer](https://github.com/DataSketches/memory/blob/master/src/main/java/com/yahoo/memory/MemoryRequestServer.java).
   It is essentially the same as the Postgres concept and has been in the 
library long before we knew about Postgres.  The underlying implementation, of 
course, would be up to Druid.  
   
   The library also has a 
[DefaultMemoryRequestServer](https://github.com/DataSketches/memory/blob/master/src/main/java/com/yahoo/memory/DefaultMemoryRequestServer.java)
 which just allocates from the heap if the primary interface has not been 
implemented.
   
   Currently, only the DS-HLL, Theta, and Quantiles sketches are designed to 
take advantage of this.  But if the demand is there, we could update other 
relevant sketches in the library to also use this mechanism.  
   
   The above MemoryRequestServer interface is pretty simple, but if it doesn't 
exactly meet your needs, let us know.  We would be happy to work with you as to 
what makes sense.
   
   
   
   

----------------------------------------------------------------
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