support Spark 3.2 and EMR 6.7 (#98)

* support Spark 3.2 and EMR 6.7 * fix previous commit format error by running black * fix missing docstring in public module * fix epel release error * fix epel release error * revert back epel change * change for CR feedback * delete setup.py
aws · Aug 31, 2022 · 3ccb729 · 3ccb729
1 parent d425af9
commit 3ccb729
Show file tree

Hide file tree

Showing 17 changed files with 1,506 additions and 12 deletions.
diff --git a/DEVELOPMENT.md b/DEVELOPMENT.md
@@ -126,14 +126,27 @@ docker push $AWS_ACCOUNT_ID.dkr.ecr.us-west-2.amazonaws.com/$SPARK_REPOSITORY:$V
 make test-sagemaker
 ```
 
+6. please run following command before you raise CR:
+
+```
+make test-unit
+make install-container-library
+```
+
+
 ## Push the code
 1. You need to create PR request in order to merge the code. How to create PR request lists here:https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/proposing-changes-to-your-work-with-pull-requests/creating-a-pull-request
 2. You need to get Github access of AWS organization. please following here:https://w.amazon.com/?Open_Source/GitHub
 3. Get access to permission specific to a team, example is here:https://github.com/orgs/aws/teams/sagemakerwrite/members
 4. Ask a person to review the code and merge it in.This repo needs at least one code reviewer. 
 5. The code needs to be signed before pushing. More detail about signing is here:https://docs.github.com/en/authentication/managing-commit-signature-verification/signing-commits.
-Remember in your local, you need to set up: git config --global user.signingkey [key id] and also upload public key into your github account.
-6. The email you specify when you created public key must match github email in github settings.
+
+```
+$ git commit -S -m "your commit message"
+```
+
+6.Remember in your local, you need to set up: git config --global user.signingkey [key id] and also upload public key into your github account.
+The email you specify when you created public key must match github email in github settings.
 
 ### FAQ
 
@@ -168,3 +181,20 @@ make: *** [install-container-library] Error 255
 ```
 
 * you need to update smsparkbuild/py39/Pipfile corresponding package version.
+
+6. Code build may fail because of the format, 
+for example
+```
+2 files would be reformatted, 13 files would be left unchanged.
+```
+
+you can fix it by running
+
+```
+black src/smspark/bootstrapper.py
+```
+see https://www.freecodecamp.org/news/auto-format-your-python-code-with-black/ for detail.
+
+7. Remember to define module at start of python file. Missing docstring error.
+
+see more detail here https://stackoverflow.com/questions/46192576/how-can-i-fix-flake8-d100-missing-docstring-error-in-atom-editor
diff --git a/Makefile b/Makefile
@@ -7,7 +7,7 @@ SHELL          := /bin/sh
 
 # Set variables if testing locally
 ifeq ($(IS_RELEASE_BUILD),)
-    SPARK_VERSION := 3.1
+    SPARK_VERSION := 3.2
     PROCESSOR := cpu
     FRAMEWORK_VERSION := py39
     SM_VERSION := 1.0

diff --git a/Pipfile b/Pipfile
@@ -0,0 +1,41 @@
+[[source]]
+name = "pypi"
+url = "https://pypi.org/simple"
+verify_ssl = true
+
+[dev-packages]
+
+[packages]
+tenacity = "==8.0.1"
+psutil = "==5.9.0"
+click = "==8.1.2"
+watchdog = "==0.10.3"
+waitress = "==2.1.2"
+types-waitress = "==2.0.6"
+requests = "==2.27.1"
+types-requests = "==2.27.16"
+rsa = "==4.3"
+pyasn1 = "==0.4.8"
+boto3 = "==1.21.33"
+safety = "==1.10.3"
+black = "==22.3.0"
+mypy = "==0.942"
+flake8 = "==4.0.1"
+flake8-docstrings = "==1.5.0"
+pytest = "==7.1.1"
+pytest-cov = "==2.10.0"
+pytest-xdist = "==2.5.0"
+docker = "==5.0.3"
+docker-compose = "==1.29.2"
+cryptography = "==36.0.2"
+typing-extensions = "==4.1.1"
+sagemaker = "==2.83.0"
+smspark = {editable = true, path = "."}
+importlib-metadata = "==4.11.3"
+pytest-parallel = "==0.1.1"
+pytest-rerunfailures = "10.0"
+numpy = "==1.22.2"
+protobuf = "==3.20.1"
+
+[requires]
+python_version = "3.9"
diff --git a/Pipfile.lock b/Pipfile.lock
diff --git a/new_images.yml b/new_images.yml
@@ -1,7 +1,7 @@
 ---
 new_images:
-  - spark: "3.1.1"
+  - spark: "3.2"
     use-case: "processing"
     processors: ["cpu"]
     python: ["py39"]
-    sm_version: "1.3"
+    sm_version: "1.0"
diff --git a/spark/processing/3.2/py3/container-bootstrap-config/bootstrap.sh b/spark/processing/3.2/py3/container-bootstrap-config/bootstrap.sh
@@ -0,0 +1 @@
+echo "Not implemented"
diff --git a/spark/processing/3.2/py3/docker/py39/Dockerfile.cpu b/spark/processing/3.2/py3/docker/py39/Dockerfile.cpu
@@ -0,0 +1,128 @@
+FROM 137112412989.dkr.ecr.us-west-2.amazonaws.com/amazonlinux:2
+ARG REGION
+ENV AWS_REGION ${REGION}
+
+RUN yum clean all \
+    && yum update -y \
+    && yum install -y awscli bigtop-utils curl gcc gzip unzip zip gunzip tar wget liblapack* libblas* libopencv* libopenblas*
+
+# Install python 3.9
+ARG PYTHON_BASE_VERSION=3.9
+ARG PYTHON_WITH_BASE_VERSION=python${PYTHON_BASE_VERSION}
+ARG PIP_WITH_BASE_VERSION=pip${PYTHON_BASE_VERSION}
+ARG PYTHON_VERSION=${PYTHON_BASE_VERSION}.12
+RUN yum -y groupinstall 'Development Tools' \
+    && yum -y install openssl-devel bzip2-devel libffi-devel sqlite-devel xz-devel \
+    && wget https://www.python.org/ftp/python/${PYTHON_VERSION}/Python-${PYTHON_VERSION}.tgz \
+    && tar xzf Python-${PYTHON_VERSION}.tgz \
+    && cd Python-*/ \
+    && ./configure --enable-optimizations \
+    && make altinstall \
+    && echo -e 'alias python3=python3.9\nalias pip3=pip3.9' >> ~/.bashrc \
+    && ln -s $(which ${PYTHON_WITH_BASE_VERSION}) /usr/local/bin/python3 \
+    && ln -s $(which ${PIP_WITH_BASE_VERSION}) /usr/local/bin/pip3 \
+    && cd .. \
+    && rm Python-${PYTHON_VERSION}.tgz \
+    && rm -rf Python-${PYTHON_VERSION}
+
+# install nginx amazonlinux:2.0.20200304.0 does not have nginx, so need to install epel-release first
+RUN wget https://dl.fedoraproject.org/pub/epel/epel-release-latest-7.noarch.rpm
+RUN yum install -y https://dl.fedoraproject.org/pub/epel/epel-release-latest-7.noarch.rpm
+RUN yum install -y nginx
+
+RUN rm -rf /var/cache/yum
+
+ENV PYTHONDONTWRITEBYTECODE=1
+ENV PYTHONUNBUFFERED=1
+# http://blog.stuart.axelbrooke.com/python-3-on-spark-return-of-the-pythonhashseed
+ENV PYTHONHASHSEED 0
+ENV PYTHONIOENCODING UTF-8
+ENV PIP_DISABLE_PIP_VERSION_CHECK 1
+
+# Install EMR Spark/Hadoop
+ENV HADOOP_HOME /usr/lib/hadoop
+ENV HADOOP_CONF_DIR /usr/lib/hadoop/etc/hadoop
+ENV SPARK_HOME /usr/lib/spark
+
+COPY yum/emr-apps.repo /etc/yum.repos.d/emr-apps.repo
+
+# Install hadoop / spark dependencies from EMR's yum repository for Spark optimizations.
+# replace placeholder with region in repository URL
+RUN sed -i "s/REGION/${AWS_REGION}/g" /etc/yum.repos.d/emr-apps.repo
+RUN adduser -N hadoop
+
+# These packages are a subset of what EMR installs in a cluster with the
+# "hadoop", "spark", and "hive" applications.
+# They include EMR-optimized libraries and extras.
+RUN yum install -y aws-hm-client \
+    aws-java-sdk \
+    aws-sagemaker-spark-sdk \
+    emr-goodies \
+    emr-ruby \
+    emr-scripts \
+    emr-s3-select \
+    emrfs \
+    hadoop \
+    hadoop-client \
+    hadoop-hdfs \
+    hadoop-hdfs-datanode \
+    hadoop-hdfs-namenode \
+    hadoop-httpfs \
+    hadoop-kms \
+    hadoop-lzo \
+    hadoop-yarn \
+    hadoop-yarn-nodemanager \
+    hadoop-yarn-proxyserver \
+    hadoop-yarn-resourcemanager \
+    hadoop-yarn-timelineserver \
+    hive \
+    hive-hcatalog \
+    hive-hcatalog-server \
+    hive-jdbc \
+    hive-server2 \
+    s3-dist-cp \
+    spark-core \
+    spark-datanucleus \
+    spark-external \
+    spark-history-server \
+    spark-python
+
+
+# Point Spark at proper python binary
+ENV PYSPARK_PYTHON=/usr/local/bin/python3.9
+
+# Setup Spark/Yarn/HDFS user as root
+ENV PATH="/usr/bin:/opt/program:${PATH}"
+ENV YARN_RESOURCEMANAGER_USER="root"
+ENV YARN_NODEMANAGER_USER="root"
+ENV HDFS_NAMENODE_USER="root"
+ENV HDFS_DATANODE_USER="root"
+ENV HDFS_SECONDARYNAMENODE_USER="root"
+
+
+
+# Set up bootstrapping program and Spark configuration
+COPY hadoop-config /opt/hadoop-config
+COPY nginx-config /opt/nginx-config
+COPY aws-config /opt/aws-config
+COPY Pipfile Pipfile.lock setup.py *.whl /opt/program/
+ENV PIPENV_PIPFILE=/opt/program/Pipfile
+# Use --system flag, so it will install all packages into the system python,
+# and not into the virtualenv. Since docker containers do not need to have virtualenvs
+# pipenv > 2022.4.8 fails to build smspark
+RUN /usr/local/bin/python3.9 -m pip install pipenv==2022.4.8 \
+    && pipenv install --system \
+    && /usr/local/bin/python3.9 -m pip install /opt/program/*.whl
+
+# Setup container bootstrapper
+COPY container-bootstrap-config /opt/container-bootstrap-config
+RUN chmod +x /opt/container-bootstrap-config/bootstrap.sh \
+    && /opt/container-bootstrap-config/bootstrap.sh
+
+# With this config, spark history server will not run as daemon, otherwise there
+# will be no server running and container will terminate immediately
+ENV SPARK_NO_DAEMONIZE TRUE
+
+WORKDIR $SPARK_HOME
+
+ENTRYPOINT ["smspark-submit"]
diff --git a/spark/processing/3.2/py3/hadoop-config/core-site.xml b/spark/processing/3.2/py3/hadoop-config/core-site.xml
@@ -0,0 +1,26 @@
+<?xml version="1.0" encoding="UTF-8"?>
+ <?xml-stylesheet type="text/xsl" href="configuration.xsl"?>
+ <!-- Put site-specific property overrides in this file. -->
+
+ <configuration>
+     <property>
+         <name>fs.defaultFS</name>
+         <value>hdfs://nn_uri/</value>
+         <description>NameNode URI</description>
+     </property>
+     <property>
+         <name>fs.s3a.aws.credentials.provider</name>
+         <value>com.amazonaws.auth.DefaultAWSCredentialsProviderChain</value>
+         <description>AWS S3 credential provider</description>
+     </property>
+     <property>
+         <name>fs.s3.impl</name>
+         <value>org.apache.hadoop.fs.s3a.S3AFileSystem</value>
+         <description>s3a filesystem implementation</description>
+     </property>
+     <property>
+         <name>fs.AbstractFileSystem.s3a.imp</name>
+         <value>org.apache.hadoop.fs.s3a.S3A</value>
+         <description>s3a filesystem implementation</description>
+     </property>
+ </configuration>
diff --git a/spark/processing/3.2/py3/hadoop-config/hdfs-site.xml b/spark/processing/3.2/py3/hadoop-config/hdfs-site.xml
@@ -0,0 +1,67 @@
+<?xml version="1.0" encoding="UTF-8"?>
+ <?xml-stylesheet type="text/xsl" href="configuration.xsl"?>
+ <!-- Put site-specific property overrides in this file. -->
+
+ <configuration>
+    <property>
+        <name>dfs.datanode.data.dir</name>
+        <value>file:///opt/amazon/hadoop/hdfs/datanode</value>
+        <description>Comma separated list of paths on the local filesystem of a DataNode where it should store its\
+  blocks.</description>
+    </property>
+
+    <property>
+        <name>dfs.namenode.name.dir</name>
+        <value>file:///opt/amazon/hadoop/hdfs/namenode</value>
+        <description>Path on the local filesystem where the NameNode stores the namespace and transaction logs per\
+ sistently.</description>
+    </property>
+
+	<!-- Fix for "Failed to replace a bad datanode on the existing pipeline due to no more good datanodes being available to try"
+		 From https://community.cloudera.com/t5/Support-Questions/Failed-to-replace-a-bad-datanode-on-the-existing-pipeline/td-p/207711
+		 This issue can be caused by Continuous network issues causing or repeated packet drops. This specially happens when data is
+		 being written to any one of the DataNode which is in process of pipelining the data to next datanode and due to any communicaiton
+		 issue it may lead to pipeline failure. We are only see this issue in small regions. -->
+	<property>
+		<name>dfs.client.block.write.replace-datanode-on-failure.enable</name>
+		<value>true</value>
+		<description>
+			If there is a datanode/network failure in the write pipeline,
+			DFSClient will try to remove the failed datanode from the pipeline
+			and then continue writing with the remaining datanodes. As a result,
+			the number of datanodes in the pipeline is decreased. The feature is
+			to add new datanodes to the pipeline.
+
+			This is a site-wide property to enable/disable the feature.
+
+			When the cluster size is extremely small, e.g. 3 nodes or less, cluster
+			administrators may want to set the policy to NEVER in the default
+			configuration file or disable this feature. Otherwise, users may
+			experience an unusually high rate of pipeline failures since it is
+			impossible to find new datanodes for replacement.
+
+			See also dfs.client.block.write.replace-datanode-on-failure.policy
+		</description>
+	</property>
+
+	<property>
+		<name>dfs.client.block.write.replace-datanode-on-failure.policy</name>
+		<value>ALWAYS</value>
+		<description>
+			This property is used only if the value of
+			dfs.client.block.write.replace-datanode-on-failure.enable is true.
+
+			ALWAYS: always add a new datanode when an existing datanode is
+			removed.
+
+			NEVER: never add a new datanode.
+
+			DEFAULT:
+			Let r be the replication number.
+			Let n be the number of existing datanodes.
+			Add a new datanode only if r is greater than or equal to 3 and either
+			(1) floor(r/2) is greater than or equal to n; or
+			(2) r is greater than n and the block is hflushed/appended.
+		</description>
+	 </property>
+ </configuration>
diff --git a/spark/processing/3.2/py3/hadoop-config/spark-defaults.conf b/spark/processing/3.2/py3/hadoop-config/spark-defaults.conf
@@ -0,0 +1,10 @@
+spark.driver.extraClassPath      /usr/lib/hadoop-lzo/lib/*:/usr/lib/hadoop/hadoop-aws.jar:/usr/share/aws/aws-java-sdk/*:/usr/share/aws/emr/emrfs/conf:/usr/share/aws/emr/emrfs/lib/*:/usr/share/aws/emr/emrfs/auxlib/*:/usr/share/aws/emr/goodies/lib/emr-spark-goodies.jar:/usr/share/aws/emr/security/conf:/usr/share/aws/emr/security/lib/*:/usr/share/aws/hmclient/lib/aws-glue-datacatalog-spark-client.jar:/usr/share/java/Hive-JSON-Serde/hive-openx-serde.jar:/usr/share/aws/sagemaker-spark-sdk/lib/sagemaker-spark-sdk.jar:/usr/share/aws/emr/s3select/lib/emr-s3-select-spark-connector.jar
+spark.driver.extraLibraryPath    /usr/lib/hadoop/lib/native:/usr/lib/hadoop-lzo/lib/native
+spark.executor.extraClassPath    /usr/lib/hadoop-lzo/lib/*:/usr/lib/hadoop/hadoop-aws.jar:/usr/share/aws/aws-java-sdk/*:/usr/share/aws/emr/emrfs/conf:/usr/share/aws/emr/emrfs/lib/*:/usr/share/aws/emr/emrfs/auxlib/*:/usr/share/aws/emr/goodies/lib/emr-spark-goodies.jar:/usr/share/aws/emr/security/conf:/usr/share/aws/emr/security/lib/*:/usr/share/aws/hmclient/lib/aws-glue-datacatalog-spark-client.jar:/usr/share/java/Hive-JSON-Serde/hive-openx-serde.jar:/usr/share/aws/sagemaker-spark-sdk/lib/sagemaker-spark-sdk.jar:/usr/share/aws/emr/s3select/lib/emr-s3-select-spark-connector.jar
+spark.executor.extraLibraryPath  /usr/lib/hadoop/lib/native:/usr/lib/hadoop-lzo/lib/native
+spark.driver.host=sd_host
+spark.hadoop.mapreduce.fileoutputcommitter.algorithm.version=2
+
+# Fix for "Uncaught exception: org.apache.spark.rpc.RpcTimeoutException: Cannot
+# receive any reply from 10.0.109.30:35219 in 120 seconds.""
+spark.rpc.askTimeout=300s
diff --git a/spark/processing/3.2/py3/hadoop-config/spark-env.sh b/spark/processing/3.2/py3/hadoop-config/spark-env.sh
@@ -0,0 +1,3 @@
+#EMPTY FILE AVOID OVERRIDDING ENV VARS
+# Specifically, without copying the empty file, SPARK_HISTORY_OPTS will be overriden, 
+# spark.history.ui.port defaults to 18082, and spark.eventLog.dir defaults to local fs
diff --git a/spark/processing/3.2/py3/hadoop-config/yarn-site.xml b/spark/processing/3.2/py3/hadoop-config/yarn-site.xml
@@ -0,0 +1,34 @@
+<?xml version="1.0"?>
+<!-- Site specific YARN configuration properties -->
+ <configuration>
+     <property>
+         <name>yarn.resourcemanager.hostname</name>
+         <value>rm_hostname</value>
+         <description>The hostname of the RM.</description>
+     </property>
+     <property>
+         <name>yarn.nodemanager.hostname</name>
+         <value>nm_hostname</value>
+         <description>The hostname of the NM.</description>
+     </property>
+     <property>
+         <name>yarn.nodemanager.webapp.address</name>
+         <value>nm_webapp_address</value>
+     </property>
+     <property>
+         <name>yarn.nodemanager.vmem-pmem-ratio</name>
+         <value>5</value>
+         <description>Ratio between virtual memory to physical memory.</description>
+     </property>
+     <property>
+         <name>yarn.resourcemanager.am.max-attempts</name>
+         <value>1</value>
+         <description>The maximum number of application attempts.</description>
+     </property>
+     <property>
+         <name>yarn.nodemanager.env-whitelist</name>
+         <value>JAVA_HOME,HADOOP_COMMON_HOME,HADOOP_HDFS_HOME,HADOOP_CONF_DIR,YARN_HOME,AWS_CONTAINER_CREDENTIALS_RELATIVE_URI,AWS_REGION</value>
+         <description>Environment variable whitelist</description>
+     </property>
+
+ </configuration>