Put in spill() and getMemorySize() implementations in HCatArrayBag
------------------------------------------------------------------
Key: HCATALOG-82
URL: https://issues.apache.org/jira/browse/HCATALOG-82
Project: HCatalog
Issue Type: Improvement
Reporter: Sushanth Sowmyan
HCatArrayBag is an implementation of the pig DataBag that can take a trivial
array of objects and expose DataBag semantics on it, so as to allow it to be
part of a Tuple, but yet not require opening and repackaging. It was intended
to be a very thin, fast implementation, but the lack of proper
spill()/getMemorySize() implementations means that Pig is unable to manage
memory efficiently when dealing with these.
Given that we're trying to push HCat to be a default storage mechanism over all
other earlier forms, we should ensure that these jagged edges are smoothed out
- we need to put in a proper implementation here.
(Note also : There are a couple of other TODOs on this - they could use fixing
too. :) )
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
