ODIn/spark-instrumented-optimizer
1
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity
spark-instrumented-optimizer/sql
History
yucai e17567ca78 [SPARK-24076][SQL] Use different seed in HashAggregate to avoid hash conflict 
## What changes were proposed in this pull request?

HashAggregate uses the same hash algorithm and seed as ShuffleExchange, it may lead to bad hash conflict when shuffle.partitions=8192*n.

Considering below example:
```
SET spark.sql.shuffle.partitions=8192;
INSERT OVERWRITE TABLE target_xxx
SELECT
 item_id,
 auct_end_dt
FROM
 from source_xxx
GROUP BY
 item_id,
 auct_end_dt;
```

In the shuffle stage, if user sets the shuffle.partition = 8192, all tuples in the same partition will meet the following relationship:
```
hash(tuple x) = hash(tuple y) + n * 8192
```
Then in the next HashAggregate stage, all tuples from the same partition need be put into a 16K BytesToBytesMap (unsafeRowAggBuffer).

Here, the HashAggregate uses the same hash algorithm on the same expression as shuffle, and uses the same seed, and 16K = 8192 * 2, so actually, all tuples in the same parititon will only be hashed to 2 different places in the BytesToBytesMap. It is bad hash conflict. With BytesToBytesMap growing, the conflict will always exist.

Before change:
<img width="334" alt="hash_conflict" src="https://user-images.githubusercontent.com/2989575/39250210-ed032d46-48d2-11e8-855a-c1afc2a0ceb5.png">

After change:
<img width="334" alt="no_hash_conflict" src="https://user-images.githubusercontent.com/2989575/39250218-f1cb89e0-48d2-11e8-9244-5a93c1e8b60d.png">

## How was this patch tested?

Unit tests and production cases.

Author: yucai <yyu1@ebay.com>

Closes #21149 from yucai/SPARK-24076.
2018-05-08 11:34:27 +02:00
..
catalyst [SPARK-24128][SQL] Mention configuration option in implicit CROSS JOIN error 2018-05-08 12:21:33 +08:00
core [SPARK-24076][SQL] Use different seed in HashAggregate to avoid hash conflict 2018-05-08 11:34:27 +02:00
hive [SPARK-24017][SQL] Refactor ExternalCatalog to be an interface 2018-05-06 20:41:32 -07:00
hive-thriftserver [SPARK-24017][SQL] Refactor ExternalCatalog to be an interface 2018-05-06 20:41:32 -07:00
create-docs.sh [MINOR][DOCS] Minor doc fixes related with doc build and uses script dir in SQL doc gen script 2017-08-26 13:56:24 +09:00
gen-sql-markdown.py [SPARK-21485][FOLLOWUP][SQL][DOCS] Describes examples and arguments separately, and note/since in SQL built-in function documentation 2017-08-05 10:10:56 -07:00
mkdocs.yml [SPARK-21485][SQL][DOCS] Spark SQL documentation generation for built-in functions 2017-07-26 09:38:51 -07:00
README.md [MINOR][DOC] Fix some typos and grammar issues 2018-04-06 13:37:08 +08:00

README.md

Spark SQL

This module provides support for executing relational queries expressed in either SQL or the DataFrame/Dataset API.

Spark SQL is broken up into four subprojects:

  • Catalyst (sql/catalyst) - An implementation-agnostic framework for manipulating trees of relational operators and expressions.
  • Execution (sql/core) - A query planner / execution engine for translating Catalyst's logical query plans into Spark RDDs. This component also includes a new public interface, SQLContext, that allows users to execute SQL or LINQ statements against existing RDDs and Parquet files.
  • Hive Support (sql/hive) - Includes an extension of SQLContext called HiveContext that allows users to write queries using a subset of HiveQL and access data from a Hive Metastore using Hive SerDes. There are also wrappers that allow users to run queries that include Hive UDFs, UDAFs, and UDTFs.
  • HiveServer and CLI support (sql/hive-thriftserver) - Includes support for the SQL CLI (bin/spark-sql) and a HiveServer2 (for JDBC/ODBC) compatible server.

Running sql/create-docs.sh generates SQL documentation for built-in functions under sql/site.