Skip to content

F
flink

doujutun3207 / flink
与 Fork 源项目一致

从无法访问的项目Fork

通知 24
Star 0
Fork 0
F
flink

体验新版 GitCode，发现更多精彩内容 >>

提交 9656e83f 编写于 作者: R Robert Metzger 提交者： Robert Metzger
浏览文件
操作

[FLINK-1234] Activate hadoop2 profile by default 

This closes #232
上级 74bc7dd3
隐藏空白更改
内联 并排
Showing with 102 addition and 114 deletion
.travis.yml
浏览文件 @ 9656e83f
......@@ -7,15 +7,15 @@ language: java
matrix:
include:
- jdk: "openjdk6"
env: PROFILE=
env: PROFILE="-Dhadoop.profile=1"
- jdk: "oraclejdk7"
env: PROFILE=
env: PROFILE="-Dhadoop.profile=1"
- jdk: "openjdk6"
env: PROFILE="-P!include-yarn -Dhadoop.profile=2 -Dhadoop.version=2.0.0-alpha"
env: PROFILE="-P!include-yarn -Dhadoop.version=2.0.0-alpha"
- jdk: "oraclejdk7"
env: PROFILE="-Dhadoop.profile=2 -Dhadoop.version=2.2.0"
env: PROFILE="-Dhadoop.version=2.2.0"
- jdk: "oraclejdk8"
env: PROFILE="-Dhadoop.profile=2 -Dhadoop.version=2.5.0"
env: PROFILE="-Dhadoop.version=2.5.1"
git:
......
docs/building.md
浏览文件 @ 9656e83f
......@@ -14,16 +14,18 @@ The simplest way of building Flink is by running:
~~~bash
cd incubator-flink
mvn clean package -DskipTests
mvn clean install -DskipTests
~~~
This instructs Maven (`mvn`) to first remove all existing builds (`clean`) and then create a new Flink binary (`package`). The `-DskipTests` command prevents Maven from executing the unit tests.
This instructs Maven (`mvn`) to first remove all existing builds (`clean`) and then create a new Flink binary (`install`). The `-DskipTests` command prevents Maven from executing the unit tests.
[Read more](http://maven.apache.org/) about Apache Maven.
## Build Flink for a specific Hadoop Version
This section covers building Flink for a specific Hadoop version. Most users do not need to do this manually.
This section covers building Flink for a specific Hadoop version. Most users do not need to do this manually. The download page of Flink contains binary packages for common setups.
The problem is that Flink uses HDFS and YARN which are both dependencies from Apache Hadoop. There exist many different versions of Hadoop (from both the upstream project and the different Hadoop distributions). If a user is using a wrong combination of versions, exceptions like this one occur:
......@@ -39,45 +41,48 @@ There are two main versions of Hadoop that we need to differentiate:
- Hadoop 2, with all versions starting with 2, like 2.2.0.
The main differentiation between Hadoop 1 and Hadoop 2 is the availability of Hadoop YARN (Hadoops cluster resource manager).
**To build Flink for Hadoop 2**, issue the following command:
By default, Flink is using the Hadoop 2 dependencies.
**To build Flink for Hadoop 1**, issue the following command:
~~~bash
mvn clean package -DskipTests -Dhadoop.profile=2
mvn clean install -DskipTests -Dhadoop.profile=1
~~~
The `-Dhadoop.profile=2` flag instructs Maven to build Flink with YARN support and the Hadoop 2 HDFS client.
The `-Dhadoop.profile=1` flag instructs Maven to build Flink for Hadoop 1. Note that the features included in Flink change when using a different Hadoop profile. In particular the support for YARN and the build-in HBase support are not available in Hadoop 1 builds.
Usually, this flag is sufficient for full support of Flink for Hadoop 2-versions.
However, you can also **specify a specific Hadoop version to build against**:
You can also **specify a specific Hadoop version to build against**:
~~~bash
mvn clean package -DskipTests -Dhadoop.profile=2 -Dhadoop.version=2.4.1
mvn clean install -DskipTests -Dhadoop.version=2.4.1
~~~
**To build Flink against a vendor specific Hadoop version**, issue the following command:
~~~bash
mvn clean package -DskipTests -Pvendor-repos -Dhadoop.profile=2 -Dhadoop.version=2.2.0-cdh5.0.0-beta-2
mvn clean install -DskipTests -Pvendor-repos -Dhadoop.version=2.2.0-cdh5.0.0-beta-2
~~~
The `-Pvendor-repos` activates a Maven [build profile](http://maven.apache.org/guides/introduction/introduction-to-profiles.html) that includes the repositories of popular Hadoop vendors such as Cloudera, Hortonworks, or MapR.
**Build Flink for `hadoop2` versions before 2.2.0**
Maven will automatically build Flink with its YARN client if the `-Dhadoop.profile=2` is set. But there were some changes in Hadoop versions before the 2.2.0 Hadoop release that are not supported by Flink's YARN client. Therefore, you can disable building the YARN client with the following string: `-P\!include-yarn`.
Maven will automatically build Flink with its YARN client. But there were some changes in Hadoop versions before the 2.2.0 Hadoop release that are not supported by Flink's YARN client. Therefore, you can disable building the YARN client with the following string: `-P\!include-yarn`.
So if you are building Flink for Hadoop `2.0.0-alpha`, use the following command:
~~~bash
-P\!include-yarn -Dhadoop.profile=2 -Dhadoop.version=2.0.0-alpha
-P\!include-yarn -Dhadoop.version=2.0.0-alpha
~~~
## Background
The builds with Maven are controlled by [properties](http://maven.apache.org/pom.html#Properties) and <a href="http://maven.apache.org/guides/introduction/introduction-to-profiles.html">build profiles</a>.
There are two profiles, one for hadoop1 and one for hadoop2. When the hadoop2 profile is enabled, the system will also build the YARN client.
The hadoop1 profile is used by default. To enable the hadoop2 profile, set `-Dhadoop.profile=2` when building.
There are two profiles, one for hadoop1 and one for hadoop2. When the hadoop2 profile is enabled (default), the system will also build the YARN client.
To enable the hadoop1 profile, set `-Dhadoop.profile=1` when building.
Depending on the profile, there are two Hadoop versions, set via properties. For "hadoop1", we use 1.2.1 by default, for "hadoop2" it is 2.2.0.
You can change these versions with the `hadoop-two.version` (or `hadoop-one.version`) property. For example `-Dhadoop-two.version=2.4.0`.
......
flink-addons/flink-hadoop-compatibility/pom.xml
浏览文件 @ 9656e83f
......@@ -66,15 +66,14 @@ under the License.
<id>hadoop-2</id>
<activation>
<property>
<!-- Please do not remove the 'hadoop1' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>hadoop.profile</name><value>2</value>
<!-- Please do not remove the 'hadoop2' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>!hadoop.profile</name>
</property>
</activation>
<dependencies>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-mapreduce-client-core</artifactId>
<version>${hadoop.version}</version>
<exclusions>
<exclusion>
<groupId>asm</groupId>
......
flink-addons/flink-yarn/pom.xml
浏览文件 @ 9656e83f
......@@ -32,7 +32,6 @@ under the License.
<packaging>jar</packaging>
<dependencies>
<dependency>
<groupId>org.apache.flink</groupId>
<artifactId>flink-runtime</artifactId>
......@@ -62,7 +61,6 @@ under the License.
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-yarn-client</artifactId>
<version>${hadoop.version}</version>
<exclusions>
<exclusion>
<groupId>asm</groupId>
......@@ -102,7 +100,6 @@ under the License.
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-common</artifactId>
<version>${hadoop.version}</version>
<exclusions>
<exclusion>
<groupId>asm</groupId>
......@@ -142,7 +139,6 @@ under the License.
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-hdfs</artifactId>
<version>${hadoop.version}</version>
<exclusions>
<exclusion>
<groupId>asm</groupId>
......@@ -182,7 +178,6 @@ under the License.
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-mapreduce-client-core</artifactId>
<version>${hadoop.version}</version>
<exclusions>
<exclusion>
<groupId>asm</groupId>
......
flink-addons/pom.xml
浏览文件 @ 9656e83f
......@@ -48,7 +48,7 @@ under the License.
<activation>
<property>
<!-- Please do not remove the 'hadoop1' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop1--><name>!hadoop.profile</name>
<!--hadoop1--><name>hadoop.profile</name><value>1</value>
</property>
</activation>
<modules>
......@@ -59,8 +59,8 @@ under the License.
<id>hadoop-2</id>
<activation>
<property>
<!-- Please do not remove the 'hadoop1' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>hadoop.profile</name><value>2</value>
<!-- Please do not remove the 'hadoop2' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>!hadoop.profile</name>
</property>
</activation>
<modules>
......@@ -71,8 +71,8 @@ under the License.
<id>include-yarn</id>
<activation>
<property>
<!-- Please do not remove the 'hadoop1' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>hadoop.profile</name><value>2</value>
<!-- Please do not remove the 'hadoop2' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>!hadoop.profile</name>
</property>
</activation>
<modules>
......
flink-dist/pom.xml
浏览文件 @ 9656e83f
......@@ -133,9 +133,12 @@ under the License.
<activation>
<property>
<!-- Please do not remove the 'hadoop1' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop1--><name>!hadoop.profile</name>
<!--hadoop1--><name>hadoop.profile</name><value>1</value>
</property>
</activation>
<properties>
<hadoop.version>${hadoop-one.version}</hadoop.version>
</properties>
<dependencies>
<!-- No extra dependencies: flink-hbase is currently not compatible with Hadoop v1 -->
</dependencies>
......@@ -145,8 +148,8 @@ under the License.
<id>include-yarn</id>
<activation>
<property>
<!-- Please do not remove the 'hadoop1' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>hadoop.profile</name><value>2</value>
<!-- Please do not remove the 'hadoop2' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>!hadoop.profile</name>
</property>
</activation>
<dependencies>
......@@ -214,8 +217,8 @@ under the License.
<id>hadoop-2</id>
<activation>
<property>
<!-- Please do not remove the 'hadoop1' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>hadoop.profile</name><value>2</value>
<!-- Please do not remove the 'hadoop2' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>!hadoop.profile</name>
</property>
</activation>
<dependencies>
......@@ -357,8 +360,6 @@ under the License.
</profile>
</profiles>
<build>
<plugins>
<plugin>
......
flink-runtime/pom.xml
浏览文件 @ 9656e83f
......@@ -178,7 +178,7 @@ under the License.
<activation>
<property>
<!-- Please do not remove the 'hadoop1' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop1--><name>!hadoop.profile</name>
<!--hadoop1--><name>hadoop.profile</name><value>1</value>
</property>
</activation>
<dependencies>
......@@ -227,8 +227,8 @@ under the License.
<id>hadoop-2</id>
<activation>
<property>
<!-- Please do not remove the 'hadoop1' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>hadoop.profile</name><value>2</value>
<!-- Please do not remove the 'hadoop2' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>!hadoop.profile</name>
</property>
</activation>
<dependencies>
......
pom.xml
浏览文件 @ 9656e83f
......@@ -79,9 +79,9 @@ under the License.
<flink.reuseForks>true</flink.reuseForks>
<slf4j.version>1.7.7</slf4j.version>
<guava.version>17.0</guava.version>
<scala.version>2.10.4</scala.version>
<scala.binary.version>2.10</scala.binary.version>
<scala.macros.version>2.0.1</scala.macros.version>
<scala.version>2.10.4</scala.version>
<scala.binary.version>2.10</scala.binary.version>
<scala.macros.version>2.0.1</scala.macros.version>
</properties>
<dependencies>
......@@ -202,7 +202,7 @@ under the License.
<version>3.2.1</version>
</dependency>
<!-- Managed dependency required for HBase in flink-hbase -->
<!-- Managed dependency required for HBase in flink-hbase -->
<dependency>
<groupId>org.javassist</groupId>
<artifactId>javassist</artifactId>
......@@ -222,6 +222,39 @@ under the License.
<artifactId>stax-api</artifactId>
<version>1.0.1</version>
</dependency>
<!-- "Old" stable Hadoop = MapReduce v1 -->
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-core</artifactId>
<version>${hadoop.version}</version>
</dependency>
<!-- YARN -->
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-common</artifactId>
<version>${hadoop.version}</version>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-hdfs</artifactId>
<version>${hadoop.version}</version>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-client</artifactId>
<version>${hadoop.version}</version>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-mapreduce-client-core</artifactId>
<version>${hadoop.version}</version>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-yarn-client</artifactId>
<version>${hadoop.version}</version>
</dependency>
</dependencies>
</dependencyManagement>
......@@ -231,54 +264,24 @@ under the License.
<activation>
<property>
<!-- Please do not remove the 'hadoop1' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop1--><name>!hadoop.profile</name>
<!--hadoop1--><name>hadoop.profile</name><value>1</value>
</property>
</activation>
<properties>
<hadoop.version>${hadoop-one.version}</hadoop.version>
</properties>
<dependencyManagement>
<dependencies>
<!-- "Old" stable Hadoop = MapReduce v1 -->
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-core</artifactId>
<version>${hadoop.version}</version>
</dependency>
</dependencies>
</dependencyManagement>
</profile>
<profile>
<id>hadoop-2</id>
<activation>
<property>
<!-- Please do not remove the 'hadoop1' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>hadoop.profile</name><value>2</value>
<!-- Please do not remove the 'hadoop2' comment. See ./tools/generate_specific_pom.sh -->
<!--hadoop2--><name>!hadoop.profile</name>
</property>
</activation>
<properties>
<hadoop.version>${hadoop-two.version}</hadoop.version>
</properties>
<dependencyManagement>
<dependencies>
<!-- YARN -->
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-common</artifactId>
<version>${hadoop.version}</version>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-hdfs</artifactId>
<version>${hadoop.version}</version>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-client</artifactId>
<version>${hadoop.version}</version>
</dependency>
</dependencies>
</dependencyManagement>
</profile>
<profile>
......@@ -498,7 +501,6 @@ under the License.
</relocations>
</configuration>
</plugin>
<plugin>
<groupId>org.apache.rat</groupId>
<artifactId>apache-rat-plugin</artifactId>
......
tools/change-version
浏览文件 @ 9656e83f
......@@ -17,4 +17,4 @@
# limitations under the License.
################################################################################
find .. -name 'pom.xml' -type f -exec sed -i 's#<version>0.7-incubating-SNAPSHOT</version>#<version>0.8-incubating-SNAPSHOT</version>#' {} \;
find .. -name 'pom.xml' -type f -exec sed -i 's#<version>0.8-hadoop2-incubating-SNAPSHOT</version>#<version>0.8-incubating-SNAPSHOT</version>#' {} \;
tools/deploy_to_maven.sh
浏览文件 @ 9656e83f
......@@ -21,12 +21,11 @@
#Please ask @rmetzger (on GitHub) before changing anything here. It contains some magic.
# Build Responsibilities
# 1. Deploy to sonatype (old hadoop)
# 2. Nothing
# 3. Deploy to s3 (old hadoop)
# 4. deploy to sonatype (yarn hadoop) (this build will also generate specific poms for yarn hadoop)
# 5. Nothing (formerly: Deploy Javadocs.)
# 6. deploy to s3 (yarn hadoop)
# 1. Deploy snapshot (hadoop1)
# 2. Deploy to s3 (hadoop1)
# 3. Nothing (hadoop200alpha)
# 4. deploy snapshot and s3 (hadoop2 - 2.2.0)
# 5. Nothing (hadoop2 - 2.5.1)
# Changes (since travis changed the id assignment)
# switched 2. with 3.
......@@ -54,12 +53,12 @@ function getVersion() {
# this will take a while
CURRENT_FLINK_VERSION=`getVersion`
if [[ "$CURRENT_FLINK_VERSION" == *-SNAPSHOT ]]; then
CURRENT_FLINK_VERSION_YARN=${CURRENT_FLINK_VERSION/-incubating-SNAPSHOT/-hadoop2-incubating-SNAPSHOT}
CURRENT_FLINK_VERSION_HADOOP1=${CURRENT_FLINK_VERSION/-incubating-SNAPSHOT/-hadoop1-incubating-SNAPSHOT}
else
CURRENT_FLINK_VERSION_YARN="$CURRENT_FLINK_VERSION-hadoop2"
CURRENT_FLINK_VERSION_HADOOP1="$CURRENT_FLINK_VERSION-hadoop1"
fi
echo "detected current version as: '$CURRENT_FLINK_VERSION' ; yarn: $CURRENT_FLINK_VERSION_YARN "
echo "detected current version as: '$CURRENT_FLINK_VERSION' ; hadoop1: $CURRENT_FLINK_VERSION_HADOOP1 "
# Check if push/commit is eligible for pushing
echo "Job: $TRAVIS_JOB_NUMBER ; isPR: $TRAVIS_PULL_REQUEST"
......@@ -71,18 +70,16 @@ if [[ $TRAVIS_PULL_REQUEST == "false" ]] ; then
#
if [[ $TRAVIS_JOB_NUMBER == *1 ]] && [[ $TRAVIS_PULL_REQUEST == "false" ]] && [[ $CURRENT_FLINK_VERSION == *SNAPSHOT* ]] ; then
# Deploy regular hadoop v1 to maven
mvn -Pdocs-and-source -DskipTests -Drat.ignoreErrors=true deploy --settings deploysettings.xml;
# Deploy hadoop v1 to maven
echo "Generating poms for hadoop1"
./tools/generate_specific_pom.sh $CURRENT_FLINK_VERSION $CURRENT_FLINK_VERSION_HADOOP1 pom.hadoop1.xml
mvn -B -f pom.hadoop1.xml -Pdocs-and-source -DskipTests -Drat.ignoreErrors=true deploy --settings deploysettings.xml;
fi
if [[ $TRAVIS_JOB_NUMBER == *4 ]] && [[ $TRAVIS_PULL_REQUEST == "false" ]] && [[ $CURRENT_FLINK_VERSION == *SNAPSHOT* ]] ; then
# deploy hadoop v2 (yarn)
echo "Generating poms for hadoop-yarn."
./tools/generate_specific_pom.sh $CURRENT_FLINK_VERSION $CURRENT_FLINK_VERSION_YARN
# all these tweaks assume a yarn build.
# performance tweaks here: no "clean deploy" so that actually nothing is being rebuild (could cause wrong poms inside the jars?)
# skip tests (they were running already)
mvn -B -f pom.hadoop2.xml -DskipTests -Pdocs-and-source -Drat.ignoreErrors=true deploy --settings deploysettings.xml;
echo "deploy standard version (hadoop2)"
mvn -B -DskipTests -Pdocs-and-source -Drat.ignoreErrors=true deploy --settings deploysettings.xml;
fi
# The block below took care of deploying javadoc to github.io. We now host the javadocs on the website.
......@@ -105,21 +102,11 @@ if [[ $TRAVIS_PULL_REQUEST == "false" ]] ; then
# cd ..
# fi
#
# Deploy binaries to S3
# The TRAVIS_JOB_NUMBER here is kinda hacked.
# Currently, there are Builds 1-6. Build 1 is deploying to maven sonatype
# Build 2 has no special meaning, it is the openjdk7, hadoop 1.2.1 build
# Build 5 is openjdk7, hadoop yarn (2.0.5-beta) build.
# Please be sure not to use Build 1 as it will always be the yarn build.
#
if [[ $TRAVIS_JOB_NUMBER == *3 ]] || [[ $TRAVIS_JOB_NUMBER == *6 ]] ; then
if [[ $TRAVIS_JOB_NUMBER == *2 ]] || [[ $TRAVIS_JOB_NUMBER == *4 ]] ; then
echo "Uploading build to amazon s3. Job Number: $TRAVIS_JOB_NUMBER"
HD="hadoop1"
# job nr 6 is YARN
if [[ $TRAVIS_JOB_NUMBER == *6 ]] ; then
# job nr 4 is YARN
if [[ $TRAVIS_JOB_NUMBER == *4 ]] ; then
# move to current dir
mkdir flink-$CURRENT_FLINK_VERSION
cp -r flink-dist/target/flink-*-bin/flink-yarn*/* flink-$CURRENT_FLINK_VERSION/
......@@ -136,7 +123,6 @@ if [[ $TRAVIS_PULL_REQUEST == "false" ]] ; then
echo "doing a ls -lisah:"
ls -lisah
fi
fi # pull request check
tools/generate_specific_pom.sh
浏览文件 @ 9656e83f
......@@ -25,7 +25,7 @@
function usage {
echo "Usage: $0 CURRENT_VERSION NEW_VERSION [POM_NAME]"
echo "For example, $0 0.4-SNAPSHOT 0.4-hadoop2-SNAPSHOT"
echo "For example, $0 0.8-incubating-SNAPSHOT 0.8-hadoop1-incubating-SNAPSHOT"
echo "Presumes VERSION has hadoop1 or hadoop2 in it. POM_NAME is optional and"
echo "allows to specify a different name for the generated pom."
exit 1
......@@ -96,7 +96,7 @@ for p in $poms; do
-e "s/\(<module>[^<]*\)/\1\/${nupom}/" \
-e "s/\(relativePath>\.\.\)/\1\/${nupom}/" \
-e "s/<!--hadoop1-->.*name>.*/${hadoop1}/" \
-e "s/<!--hadoop2-->.*<name>.*/${hadoop2}/" \
-e "s/<!--hadoop2-->.*name>.*/${hadoop2}/" \
$p > "$tmp_nuname"
mv $tmp_nuname $nuname
done
......
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或