before beginning to scan a database, all the executors need a licence file. I have a function licenceFileInstaller.install() that downloads the file at the right location on the system, which works fine on the master node. How do I run this on all the executors ?
I guess when you use this method in rdd transformations, it should serialize this method and send it across all the nodes. Like you can use this method within mapPartitions and it will be distributed across all the nodes and run once for every partition.
rdd.mapPartitions( x => ( licenceFileInstaller.install() ...further database code ))
Also, I guess you can add that license file using sc.addFile() which will be distributed across all the nodes.
You can refer to Spark closures for further understanding
Related
I'm trying to write a batch job to process a couple of hundreds of terabytes that currently sit in an HBase database (in an EMR cluster in AWS), all in a single large table. For every row I'm processing, I need to get additional data from a lookup table (a simple integer to string mapping) that is in a second HBase table. We'd be doing 5-10 lookups per row.
My current implementation uses a Spark job that's distributing partitions of the input table to its workers, in the following shape:
Configuration hBaseConfig = newHBaseConfig();
hBaseConfig.set(TableInputFormat.SCAN, convertScanToString(scan));
hBaseConfig.set(TableInputFormat.INPUT_TABLE, tableName);
JavaPairRDD<ImmutableBytesWritable, Result> table = sparkContext.newAPIHadoopRDD(hBaseConfig, TableInputFormat.class, ImmutableBytesWritable.class, Result.class);
table.map(val -> {
// some preprocessing
}).foreachPartition(p -> {
p.forEachRemaining(row -> {
// code that does the lookup
});
});
The problem is that the lookup table is too big to fit in the workers' memory. They all need access to all parts of the lookup table, but their access pattern would significantly benefit from a cache.
Am I right in thinking that I cannot use a simple map as a broadcast variable because it'd need to fit into memory?
Spark uses a shared nothing architecture, so I imagine there won't be an easy way to share a cache across all workers, but can we build a simple LRU cache for every individual worker?
How would I implement such a local worker cache that gets the data from the lookup table in HBase on a cache miss? Can I somehow distribute a reference to the second table to all workers?
I'm not set on my choice of technology, apart from HBase as the data source. Is there a framework other than Spark which could be a better fit for my use case?
You have a few of options for dealing with this requirement:
1- Use RDD or Dataset joins
You can load both of your HBase tables as Spark RDD or Datasets and then do a join on your lookup key.
Spark will split both RDD into partitions and shuffle content around so that rows with the same keys end up on the same executors.
By managing the number of number of partitions within spark you should be able to join 2 tables on any arbitrary sizes.
2- Broadcast a resolver instance
Instead of broadcasting a map, you can broadcast a resolver instance that does a HBase lookup and temporary LRU cache. Each executor will get a copy of this instance and can manage its own cache and you can invoke them within for foreachPartition() code.
Beware, the resolver instance needs to implement Serializable so you will have to declare the cache, HBase connections and HBase Configuration properties as transient to be initialized on each executor.
I run such a setup in Scala on one of the projects I maintain: it works and can be more efficient than the straight Spark join if you know your access patterns and manage you cache efficiently
3- Use the HBase Spark connector to implement your lookup logic
Apache HBase has recently incorporated improved HBase Spark connectors
The documentation is pretty sparse right now, you need to look at the JIRA tickets and the documentation of the previous incarnation of these tools
Cloudera's SparkOnHBase but the last unit test in the test suite looks pretty much like what you want
I have no experience with this API though.
I am currently running some Spark code and I need to query a data frame that is taking a long time (over 1 hour) per query. I need to query multiple times to check if the data frame is in fact correct.
I am relatively new to Spark and I understand that Spark uses lazy evaluation which means that the commands are executed only once I do a call for some action (in my case .show()).
Is there a way to do this process once for the whole DF and then quickly call on the data?
Currently I am saving the DF as a temporary table and then running queries in beeline (HIVE). This seems a little bit overkill as I have to save the table in a database first, which seems like a waste of time.
I have looked into the following functions .persist, .collect but I am confused on how to use them and query from them.
I would really like to learn the correct way of doing this.
Many thanks for the help in advance!!
Yes, you can keep your RDD in memory using rddName.cache() (or persists()) . More information about RDD Persistence can be found here
Using a temporary table ( registerTempTable (spark 1.6) or createOrReplaceTempView (spark2.x)) does not "save" any data. It only creates a view with the lifetime of you spark session. If you wish to save the table, you should use .saveAsTable, but I assume that this is not what you are looking for.
Using .cache is equivalent to .persist(StorageLevel.MEMORY). If your table is large and thus can't fit in memory, you should use .persist(StorageLevel.MEMORY_AND_DISK).
Also it is possible that you simple need more nodes in you cluster. In case you are running locally, make sure you deploy with --master local[*] to use all available cores on your machine. If you are running on a stand alone cluster or with a cluster manager like Yarn or Mesos, you should make sure that all necessary/available resources are assigned to you job.
I don't have any RDD to use, I just want to execute some of my own functions on some nodes of my cluster, with Apache Spark. So I don't have any data to distribute, but only code (which depends on the node that is executing it).
Is it possible ? Is Spark compatible with this goal ?
Is it possible?
I think it is possible and I've been asked about it few times already (so had time to think about it :))
Is Spark compatible with this goal?
The way Spark could handle it is to launch as many executors as you want to use nodes for the distributed work. That's the job of a cluster manager to spread the work across a cluster of nodes and so Spark can only use what nodes are given.
With the nodes assigned you simply execute a computation on fake dataset to build a RDD on top of.
If the computation runs on a node that should not be used, you can hostname inside the code and see what node you are on and decide on whether to continue or stop.
You could even read the code to execute from a database (seen a solution like this already).
I am new to distributed computing, and I'm trying to run Kmeans on EC2 using Spark's mllib kmeans. As I was reading through the tutorial I found the following code snippet on
http://spark.apache.org/docs/latest/mllib-clustering.html#k-means
I am having trouble understanding how this code runs inside the cluster. Specifically, I'm having trouble understanding the following:
After submitting the code to master node, how does spark know how to parallelize the job? Because there seem to be no part of the code that deals with this.
Is the code copied to all nodes and executed on each node? Does the master node do computation?
How do node communitate the partial result of each iteration? Is this dealt inside the kmeans.train code, or is the spark core takes care of it automatically?
Spark divides data to many partitions. For example, if you read a file from HDFS, then partitions should be equal to partitioning of data in HDFS. You can manually specify number of partitions by doing repartition(numberOfPartitions). Each partition can be processed on separate node, thread, etc. Sometimes data are partitioned by i.e. HashPartitioner, which looks on hash of the data.
Number of partitions and size of partitions generally tells you if data is distributed/parallelized correctly. Creating partitions of data is hidden in RDD.getPartitions methods.
Resource scheduling depends on cluster manager. We can post very long post about them ;) I think that in this question, the partitioning is the most important. If not, please inform me, I will edit answer.
Spark serializes clusures, that are given as arguments to transformations and actions. Spark creates DAG, which is sent to all executors and executors execute this DAG on the data - it launches closures on each partition.
Currently after each iteration, data is returned to the driver and then next job is scheduled. In Drizzle project, AMPLab/RISELab is creating possibility to create multiple jobs on one time, so data won't be sent to the driver. It will create DAG one time and schedules i.e. job with 10 iterations. Shuffle between them will be limited / will not exists at all. Currently DAG is created in each iteration and job in scheduled to executors
There is very helpful presentation about resource scheduling in Spark and Spark Drizzle.
I'm trying to benchmark a few approaches to putting an image processing algorithm into apache spark. For one step in this algorithm, a computation on a pixel in the image will depend on an unknown amount of surrounding data, so we can't partition the image with guaranteed sufficient overlap a priori.
One solution to that problem I need to benchmark is for a worker node to ask the master node for more data when it encounters a pixel with insufficient surrounding data. I'm not convinced this is the way to do things, but I need to benchmark it anyway because of reasons.
Unfortunately, after a bunch of googling and reading docs I can't find any way for a processingFunc called as part of sc.parallelize(partitions).map(processingFunc) to query the master node for more data from a different partition mid-computation.
Does a way for a worker node to ask the master for more data exist in spark, or will I need to hack something together that kind of goes around spark?
Master Node in Spark is for allocating the resources to a particular job and once the resources are allocated, the Driver ships the complete code with all its dependencies to the various executors.
The first step in every code is to load the data to the Spark cluster. You can read the data from any underlying data repository like Database, filesystem, webservices etc.
Once data is loaded it is wrapped into an RDD which is partitioned across the nodes in the cluster and further stored in the workers/ Executors Memory. Though you can control the number of partitions by leveraging various RDD API's but you should do it only when you have valid reasons to do so.
Now all operations are performed over RDD's using its various methods/ Operations exposed by RDD API. RDD keep tracks of partitions and partitioned data and depending upon the need or request it automatically query the appropriate partition.
In nutshell, you do not have to worry about the way data is partitioned by RDD or which partition stores which data and how they communicate with each other but if you do care, then you can write your own custom partitioner, instructing Spark of how to partition your data.
Secondly if your data cannot be partitioned then I do not think Spark would be an ideal choice because that will result in processing of everything in 1 single machine which itself is contrary to the idea of distributed computing.
Not sure what is exactly your use case but there are people who have been leveraging Spark for Image processing. see here for the comments from Databricks