A highly configurable Google Cloud Dataflow pipeline that writes data into a Google Big Query table from Pub/Sub
Compile this Dataflow pipeline into a runnable Jar (pubsub-to-bq.jar). Then call the jar with following parameters:
java.exe -jar "C:\Jars\pubsub-to-bq.jar" --runner=BlockingDataflowPipelineRunner --params="<params><workingBucket>gs://your_bucket</workingBucket><maxNumWorkers>1</maxNumWorkers><diskSizeGb>250</diskSizeGb><machineType>n1-standard-1</machineType><keyFile>C:\KeyFiles\YourFile.json</keyFile><accountEmail>your_account@developer.gserviceaccount.com</accountEmail><projectId>your_project_id</projectId><pipelineName>your_pipeline_name</pipelineName><pubSubTopic>your_pub_topic</pubSubTopic><bqDataSet>your_destination_BQ_dataset</bqDataSet><bqTable>your_destination_BQ_table</bqTable><streaming>true</streaming><zone>us-west1-a</zone><schema>{"fields":[{"description":null,"fields":null,"mode":"REQUIRED","name":"Student_Name","type":"STRING","ETag":null}],"ETag":null}</schema></params>"Parameters formatted view:
<params>
<workingBucket>gs://your_bucket</workingBucket>
<maxNumWorkers>1</maxNumWorkers>
<diskSizeGb>250</diskSizeGb>
<machineType>n1-standard-1</machineType>
<keyFile>C:\KeyFiles\YourFile.json</keyFile>
<accountEmail>your_account@developer.gserviceaccount.com</accountEmail>
<projectId>your_project_id</projectId>
<pipelineName>your_pipeline_name</pipelineName>
<pubSubTopic>your_pub_topic</pubSubTopic>
<bqDataSet>your_destination_BQ_dataset</bqDataSet>
<bqTable>your_destination_BQ_table</bqTable>
<streaming>true</streaming>
<zone>us-west1-a</zone>
<schema>{"fields":[{"description":null,"fields":null,"mode":"REQUIRED","name":"Student_Name","type":"STRING","ETag":null}],"ETag":null}</schema>
</params>- Only provide .json key files for GCP.
- Pipeline can support queues with batched messages.
- Set the debug flag setting if you want pipeline to log incoming messages
- This pipeline can also support streaming data into a date partitioned BQ table with a streaming pipeline. Details