Adding dockerfile

Author: pravinbhat

Dockerfile

@@ -0,0 +1,24 @@
+FROM eclipse-temurin:8-jammy
+
+RUN apt update && apt install -y openssh-server vim
+RUN service ssh start
+
+RUN mkdir -p /assets/
+
+# Build the jar locally before building docker image
+COPY ./target/cassandra-data-migrator-2.9.jar /assets/
+
+COPY ./src/resources/sparkConf.properties /assets/
+COPY ./src/resources/partitions.csv /assets/
+COPY ./src/resources/primary_key_rows.csv /assets/
+COPY ./src/resources/runCommands.txt /assets/
+
+RUN cd /assets && curl -OL https://downloads.datastax.com/dsbulk/dsbulk.tar.gz && tar -xzf ./dsbulk.tar.gz && rm ./dsbulk.tar.gz
+RUN cd /assets && curl -OL https://downloads.datastax.com/enterprise/cqlsh-astra.tar.gz && tar -xzf ./cqlsh-astra.tar.gz && rm ./cqlsh-astra.tar.gz
+RUN cd /assets && curl -OL https://archive.apache.org/dist/spark/spark-2.4.8/spark-2.4.8-bin-hadoop2.7.tgz && tar -xzf ./spark-2.4.8-bin-hadoop2.7.tgz && rm ./spark-2.4.8-bin-hadoop2.7.tgz
+
+ENV PATH="${PATH}:/assets/dsbulk-1.10.0/bin/:/assets/cqlsh-astra/bin/:/assets/spark-2.4.8-bin-hadoop2.7/bin/"
+
+EXPOSE 22
+
+CMD ["/usr/sbin/sshd","-D"]

pom.xml

@@ -3,7 +3,7 @@
 
   <groupId>datastax.astra.migrate</groupId>
   <artifactId>cassandra-data-migrator</artifactId>
-  <version>2.9</version>
+  <version>2.10</version>
   <packaging>jar</packaging>
 
   <properties>

src/resources/runCommands.txt

@@ -1,14 +1,14 @@
-// Download spark
-wget https://archive.apache.org/dist/spark/spark-2.4.8/spark-2.4.8-bin-hadoop2.6.tgz
-
-// Increase driver memory
---driver-memory 8G
+// Download dependencies
+curl -OL https://downloads.datastax.com/dsbulk/dsbulk.tar.gz
+curl -OL https://downloads.datastax.com/enterprise/cqlsh-astra.tar.gz
+wget https://archive.apache.org/dist/spark/spark-2.4.8/spark-2.4.8-bin-hadoop2.7.tgz
 
 // Migrate
-spark-submit --properties-file /<path>/sparkConf.properties --verbose --master "local[8]" --class datastax.astra.migrate.Migrate /<path>/cassandra-data-migrator-1.x.jar
+spark-submit --properties-file /<path>/sparkConf.properties --verbose --master "local[8]" --class datastax.astra.migrate.Migrate /<path>/cassandra-data-migrator-2.x.jar
+spark-submit --properties-file /<path>/sparkConf.properties --master "local[8]" --driver-memory 25G --executor-memory 25G --class datastax.astra.migrate.Migrate /<path>/cassandra-data-migrator-2.x.jar &> table_out.log
 
 // Random Partitioner Run Command
-spark-submit --properties-file /<path>/sparkConf.properties --verbose --master "local[8]" --conf spark.source.minPartition=-1 --conf spark.source.maxPartition=170141183460469231731687303715884105728 --class datastax.astra.migrate.Migrate /<path>/cassandra-data-migrator-1.x.jar
+spark-submit --properties-file /<path>/sparkConf.properties --verbose --master "local[8]" --conf spark.origin.minPartition=-1 --conf spark.origin.maxPartition=170141183460469231731687303715884105728 --class datastax.astra.migrate.Migrate /<path>/cassandra-data-migrator-1.x.jar
 
-// Diff Data
-spark-submit --properties-file /<path>/sparkConf.properties --verbose --master "local[8]" --class datastax.astra.migrate.DiffData /<path>/cassandra-data-migrator-1.x.jar
+// Validate
+spark-submit --properties-file /<path>/sparkConf.properties --master "local[8]" --driver-memory 25G --executor-memory 25G --class datastax.astra.migrate.DiffData /<path>/cassandra-data-migrator-2.x.jar &> table_out.log