aboutsummaryrefslogtreecommitdiffstats
path: root/README.md
diff options
context:
space:
mode:
authorcra14 <chandan.rajah2@bskyb.com>2013-09-04 14:00:44 +0100
committercra14 <chandan.rajah2@bskyb.com>2013-09-04 14:00:44 +0100
commitcefd1a9332d674f0ccf69ab1c7bdc102c2a48c92 (patch)
treeef02d8cd1cd14dec1988eb3765a8ff4f892a8922 /README.md
parent3501e241a2313cf49c371630cb6ebe0c3a47e991 (diff)
downloadSpyGlass-cefd1a9332d674f0ccf69ab1c7bdc102c2a48c92.tar.gz
SpyGlass-cefd1a9332d674f0ccf69ab1c7bdc102c2a48c92.zip
Moved maven repo locally
Added documentation. Scala 2.9.3 and 2.10.2 can be toggled
Diffstat (limited to 'README.md')
-rw-r--r--README.md75
1 files changed, 73 insertions, 2 deletions
diff --git a/README.md b/README.md
index beaa4a4..6df2fb8 100644
--- a/README.md
+++ b/README.md
@@ -5,6 +5,13 @@ Cascading and Scalding wrapper for HBase with advanced read and write features.
Prevent Hot Spotting by the use of transparent key prefixes.
+Changes
+=======
+- Added JDBC Tap Functionality
+- Added Delete Functionality
+- Added Region grouping of splits
+- Migrated maven repo to local subdirectory. See below.
+
Building
========
@@ -17,7 +24,7 @@ To use SpyGlass as a dependency use the following repository
<repositories>
<repository>
<id>parallelai-releases</id>
- <url>https://github.com/ParallelAI/mvn-repo/raw/master/releases</url>
+ <url>https://github.com/ParallelAI/SpyGlass/raw/master/releases</url>
</repository>
</repositories>
@@ -270,4 +277,68 @@ e.g.
6. Jdbc Tap and Source
===========================
-To be added soon. \ No newline at end of file
+To be added soon.
+
+e.g.
+ val jdbcSourceRead = new JDBCSource(
+ "TABLE_01",
+ "com.mysql.jdbc.Driver",
+ "jdbc:mysql://localhost:3306/sky_db?zeroDateTimeBehavior=convertToNull",
+ "root",
+ "password",
+ List("ID", "TEST_COLUMN1", "TEST_COLUMN2", "TEST_COLUMN3"),
+ List("bigint(20)", "varchar(45)", "varchar(45)", "bigint(20)"),
+ List("id"),
+ new Fields("key", "column1", "column2", "column3"),
+ null, null, null
+ )
+
+ val jdbcSourceWrite = new JDBCSource(
+ "TABLE_01",
+ "com.mysql.jdbc.Driver",
+ "jdbc:mysql://localhost:3306/sky_db?zeroDateTimeBehavior=convertToNull",
+ "root",
+ "password",
+ List("ID", "TEST_COLUMN1", "TEST_COLUMN2", "TEST_COLUMN3"),
+ List("bigint(20)", "varchar(45)", "varchar(45)", "bigint(20)"),
+ List("id"),
+ new Fields("key", "column1", "column2", "column3"),
+ null, null, null
+ )
+
+
+7. HBase Delete Functionality
+=============================
+
+Delete functionality has been added to SpyGlass version 4.1.0 onwards. Below is an example of how to use it.
+
+The feature can be enable by using the parameter sinkMode = SinkMode.REPLACE in the HBaseSource
+You can use sinkMode = SinkMode.UPDATE to add to the HBaseTable as usual.
+
+e.g.
+ val eraser = toIBW(input, TABLE_SCHEMA)
+ .write(new HBaseSource( "_TEST.SALT.01", quorum, 'key,
+ TABLE_SCHEMA.tail.map((x: Symbol) => "data"),
+ TABLE_SCHEMA.tail.map((x: Symbol) => new Fields(x.name)), sinkMode = SinkMode.REPLACE ))
+
+All rows with the key will be deleted. This includes all versions too.
+
+8. Regional Split Group in Scan and Get
+=======================================
+
+This functionality reduces the number of mappers significantly when using hot spot prevention with prefixes.
+
+This feature can be activated by using inputSplitType = SplitType.REGIONAL
+You can use inputSplitType = SplitType.GRANULAR to use the previous functionality as is.
+
+e.g.
+ val hbase04 = new HBaseSource( "_TEST.SALT.01", quorum, 'key,
+ TABLE_SCHEMA.tail.map((x: Symbol) => "data"),
+ TABLE_SCHEMA.tail.map((x: Symbol) => new Fields(x.name)),
+ sourceMode = SourceMode.SCAN_RANGE, startKey = sttKeyP, stopKey = stpKeyP,
+ inputSplitType = splitType).read
+ .fromBytesWritable(TABLE_SCHEMA )
+ .map(('key, 'salted, 'unsalted) -> 'testData) {x: (String, String, String) => List(x._1, x._2, x._3)}
+ .project('testData)
+ .write(TextLine("saltTesting/ScanRangeNoSalt01"))
+ .groupAll(group => group.toList[List[List[String]]]('testData -> 'testData)) \ No newline at end of file