TAJO-907: Implement off-heap tuple block and zero-copy tuple.#133
TAJO-907: Implement off-heap tuple block and zero-copy tuple.#133hyunsik wants to merge 64 commits intoapache:masterfrom
Conversation
…into DirectMemTuple
…into OffHeapTuple
…into OffHeapTuple
|
As I mentioned, I improved ExternalSort to make use of OffHeapRowBlock. I'd like to share a brief benchmark test result. Experimental environment was as follows:
Data set is lineitem table of TPC-H 1GB. The test query is Current implementation: 55 - 60 seconds Roughly, it reduces about 50% query response time. This change only affects physical operators, so I think that a local cluster benchmark is sufficient. Later, I'll adopt OffHeapRowBlock to other operators and our operator model sequentially. |
|
Rebased and implemented clone() of HeapTuple and UnSafeTuple. |
…into TAJO-907 Conflicts: tajo-core/src/main/java/org/apache/tajo/worker/TajoWorker.java tajo-core/src/main/java/org/apache/tajo/worker/Task.java
|
@hyunsik I can't build on JDK 1.6 java version : 1.6.0_65 |
|
I've removed the constant variables of 1.7, and rebased it. |
|
Committed it to block_iteration branch. |
Ability to use zeppelin with ssl both https and wss
Hi folks,
One week ago, I contributed run-time code generation for computation efficiency and reduction of objection creations. This work has also the similar purpose.
This patch includes off-heap memory and row block (a list of tuple) using off-heap memory container. Also, it provides various utility classes for them.
In detail, this patch includes lots of changes.
Later, I'll replace current VTuple by HeapTuple or ZeroCopyTuple. Also, I'm planning the improvement to replace current pull-iterator model of execution engine by the push-based block iterator model. I'll describe it later in another jira issue.
Thanks,
Hyunsik