Apache Spark (Structured Streaming) : S3 Checkpoint support

Question

From the spark structured streaming documentation: "This checkpoint location has to be a path in an HDFS compatible file system, and can be set as an option in the DataStreamWriter when starting a query."

And sure enough, setting the checkpoint to a s3 path throws:

17/01/31 21:23:56 ERROR ApplicationMaster: User class threw exception: java.lang.IllegalArgumentException: Wrong FS: s3://xxxx/fact_checkpoints/metadata, expected: hdfs://xxxx:8020 
java.lang.IllegalArgumentException: Wrong FS: s3://xxxx/fact_checkpoints/metadata, expected: hdfs://xxxx:8020 
        at org.apache.hadoop.fs.FileSystem.checkPath(FileSystem.java:652) 
        at org.apache.hadoop.hdfs.DistributedFileSystem.getPathName(DistributedFileSystem.java:194) 
        at org.apache.hadoop.hdfs.DistributedFileSystem.access$000(DistributedFileSystem.java:106) 
        at org.apache.hadoop.hdfs.DistributedFileSystem$22.doCall(DistributedFileSystem.java:1305) 
        at org.apache.hadoop.hdfs.DistributedFileSystem$22.doCall(DistributedFileSystem.java:1301) 
        at org.apache.hadoop.fs.FileSystemLinkResolver.resolve(FileSystemLinkResolver.java:81) 
        at org.apache.hadoop.hdfs.DistributedFileSystem.getFileStatus(DistributedFileSystem.java:1301) 
        at org.apache.hadoop.fs.FileSystem.exists(FileSystem.java:1430) 
        at org.apache.spark.sql.execution.streaming.StreamMetadata$.read(StreamMetadata.scala:51) 
        at org.apache.spark.sql.execution.streaming.StreamExecution.<init>(StreamExecution.scala:100) 
        at org.apache.spark.sql.streaming.StreamingQueryManager.createQuery(StreamingQueryManager.scala:232) 
        at org.apache.spark.sql.streaming.StreamingQueryManager.startQuery(StreamingQueryManager.scala:269) 
        at org.apache.spark.sql.streaming.DataStreamWriter.start(DataStreamWriter.scala:262) 
        at com.roku.dea.spark.streaming.FactDeviceLogsProcessor$.main(FactDeviceLogsProcessor.scala:133) 
        at com.roku.dea.spark.streaming.FactDeviceLogsProcessor.main(FactDeviceLogsProcessor.scala) 
        at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) 
        at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) 
        at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) 
        at java.lang.reflect.Method.invoke(Method.java:498) 
        at org.apache.spark.deploy.yarn.ApplicationMaster$$anon$2.run(ApplicationMaster.scala:637) 
17/01/31 21:23:56 INFO SparkContext: Invoking stop() from shutdown hook 

A couple of questions here:

1. Why is s3 not supported as a checkpoint dir (regular spark streaming supports this)? What makes a filesystem "HDFS compliant" ?
2. I use HDFS emphemerally (since clusters can come up or down all the time) and use s3 as the place to persist all data - what would be the recommendations for storing checkpointing data for structured streaming data in such a setup?

Answer 1

What makes an FS HDFS "compliant?" it's a file system, with the behaviours specified in Hadoop FS specification. The difference between an object store and FS is covered there, with the key point being "eventually consistent object stores without append or O(1) atomic renames are not compliant"

For S3 in particular

1. It's not consistent: after a new blob is created, a list command often doesn't show it. Same for deletions.
2. When a blob is overwritten or deleted, it can take a while to go away
3. rename() is implemented by copy and then delete

Spark streaming checkpoints by saving everything to a location and then renaming it to the checkpoint directory. This makes the time to checkpoint proportional to the time to do a copy of the data in S3, which is ~6-10 MB/s.

The current bit of streaming code isn't suited for s3

For now, do one of

• checkpoint to HDFS and then copy over the results
• checkpoint to a bit of EBS allocated and attached to your cluster
• checkpoint to S3, but have a long gap between checkpoints so that the time to checkpoint doesn't bring your streaming app down.

If you are using EMR, you can pay the premium for a consistent, dynamo DB backed S3, which gives you better consistency. But copy time is still the same, so checkpointing will be just as slow

Answer 2

This is a known issue: https://issues.apache.org/jira/browse/SPARK-19407

Should be fixed in the next release. You can set the default file system to s3 using --conf spark.hadoop.fs.defaultFS=s3 as a workaround.

Answer 3

This problem is fixed in https://issues.apache.org/jira/browse/SPARK-19407.

However Structured Streaming checkpointing doesn't work well in S3 because of lack of eventual consistency in S3. It's not a good idea to use S3 for checkpointing https://issues.apache.org/jira/browse/SPARK-19013.

Micheal Armburst has said that this won't be fixed in Spark, and the solution is to wait for S3guard to be implemented. S3Guard is sometime away.

Answer 4

Yes, if you are using Spark Structured Streaming version 3 or above. First, create a SparkSession and add S3 configs to its context.

val sparkSession = SparkSession
    .builder()
    .master(sparkMasterUrl)
    .appName(appName)
    .getOrCreate()

sparkSession.sparkContext.hadoopConfiguration.set("fs.s3a.access.key", "accessKey")
sparkSession.sparkContext.hadoopConfiguration.set("fs.s3a.secret.key", "secretKey")
sparkSession.sparkContext.hadoopConfiguration.set("fs.s3a.endpoint", "http://s3URL:s3Port")
sparkSession.sparkContext.hadoopConfiguration.set("spark.hadoop.fs.s3a.impl", "org.apache.hadoop.fs.s3a.S3AFileSystem")

Later, add checkpointLocation config with S3 bucket before you start the query. For example:

val streamingQuery = streamingDF.writeStream
    .option("checkpointLocation", "s3a://bucketName/checkpointDir/")
    .foreachBatch{(batchDF: DataFrame, batchId: Long) =>
       // Transform and write batchDF
     }.start()

streamingQuery.awaitTermination()

Answer 5

you can use s3 for checkpoint but you should enable EMRFS, so that s3 consistency will be handled.