8 SQL-on-Hadoop frameworks worth checking out

The language of data is SQL, so naturally lots of tools have been developed to bring SQL to Hadoop. They range from simple wrappers on top of Map Reduce to full data warehouse implementations built on top of HDFS and everywhere in between.

There are more tools than you might think, so this is my attempt at listing them all and hopefully providing some insight into what each of them actually does.

I’ve tried to order them based on ‘installation friction’, so the more complex products are towards the bottom.

I’ll cover the following technologies:

• Apache Hive
• Impala
• Presto (Facebook)
• Shark
• Apache Drill
• EMC/Pivotal HAWQ
• BigSQL by IBM
• Apache Pheonix (for HBase)
• Apache Tajo

Apache Hive

Hive is the original SQL-on-Hadoop solution.

Hive is an open-source Java project which converts SQL to a series of Map-Reduce jobs which run on standard Hadoop tasktrackers. It tries to look like MySQL by using a metastore (itself a database) to store table schemas, partitions, and locations. It largely supports MySQL syntax and organizes datasets using familiar database/table/viewconventions. Hive provides:

• A SQL-like query interface called Hive-QL, loosely modelled after MySQL
• A command line client
• Metadata sharing via a central service
• JDBC drivers
• Multi-language Apache Thrift drivers
• A Java API for creating custom functions and transformations

SHOULD YOU USE IT?

Hive is considered one of the de-facto tools installed on almost all Hadoop installations. It’s simple to set up and doesn’t require much infrastructure to get started with. Given the small cost of use, there’s pretty much no reason to not try it.

That said, queries performed with Hive are usually very slow because of the overhead associated with using Map-Reduce.

THE FUTURE OF HIVE

Hortonworks has been pushing the development of Apache Tez as a new back-end for Hive to provide fast response times currently unachievable using Map Reduce.

HIVE TUTORIAL VIDEO

Cloudera Impala

Impala is an open-source ‘interactive’ SQL query engine for Hadoop. It’s build byCloudera, one of the largest Hadoop vendors in the market. Like Hive, it provides a way to write SQL queries against your existing Hadoop data. Unlike Hive, it does not use Map-Reduce to execute the queries, but instead uses it’s own set of execution daemons which need to be installed alongside your datanodes. It provides:

• ANSI-92 SQL syntax support
• HIVE-QL support
• A command line client
• ODBC drivers
• Interop with the Hive Metastore for schema sharing across platforms
• A C++ api for creating functions and transformations

SHOULD YOU USE IT?

Impala is designed to complement the use of Apache Hive, so if you need faster access to your data than Hive can offer it might be a good choice, especially if you have a Cloudera, MapR or Amazon Hadoop cluster deployed. That said, to fully benefit from Impala’s architecture you’ll need to store your data in a particular file format (Parquet), which can be a painful transition. Additionally, you’ll need to install Impala deamons across your cluster which means taking some resources away from your Tasktrackers. Impala does not currently support YARN.

THE FUTURE OF IMPALA

Cloudera has been prototyping Impala integration for YARN which could make the deployment of Impala much less painful for the next generation of Hadoop clusters.

IMPALA VIDEO

 

Presto by Facebook

Presto is an open source ‘interactive’ SQL query engine for Hadoop written in Java. It’s built by Facebook, the original creators of Hive. Presto is similar in approach to Impala in that it is designed to provide an interactive experience whilst still using your existing datasets stored in Hadoop. It also requires installation across many ‘nodes’, again similar to Impala. It provides:

• ANSI-SQL syntax support (presumably ANSI-92)
• JDBC Drivers
• A set of ‘connectors’ used to read data from existing data sources. Connectors include: HDFS, Hive, and Cassandra.
• Interop with the Hive metastore for schema sharing

SHOULD YOU USE IT?

Presto targets the same goals as Cloudera’s Impala. Unlike Impala, it is not supported by a major vendor, so if you need enterprise support for your installation you’re out of luck. That said, it is in production use by some well-known and respected technology companies, so presumably there is somewhat of a community around it. Like Impala, performance is dependent on storing data in a particular format (RCFile). Honestly I’d think carefully about your abilities to support and debug Presto before deploying it, if you’re comfortable with those aspects of it, and you trust Facebook not to abandon the open source version of Presto then go for it.

PRESTO VIDEO

 

Shark

Shark is an open source SQL query engine written in Scala by UC Berkeley. Like Impala and Presto, it is designed to complement an existing Hive installation, and executes queries on it’s own set of worker nodes instead of using Map-Reduce. Unlike Impala and Presto, Shark is built on top of an existing data processing engine called Apache Spark. Spark is very popular right now, and is starting to build quite a large community. Think of Spark as a faster alternative to Map-Reduce. Shark provides:

• SQL-like query language supporting most of Hive-QL
• A command line client (basically the Hive client)
• Interoperability with the Hive metastore for schema sharing
• Support for existing Hive extensions such as UDFs and SerDes

SHOULD YOU USE IT?

Shark is interesting because it aims to support the entirity of Hive functionality whilst also trying to offer massive performance improvements. That said, while a lot of organizations are using Spark, I’m not sure how many are using Shark. I don’t think it provides the same sort of performance improvements offered by Presto and Impala, but if you already plan on using Spark it seems like a no-brainer to at least try it, especially as Spark is being supported by a lot of major vendors.

SHARK VIDEO

 

Apache Drill

Oh boy, another one. Apache Drill is an open-source ‘interactive’ SQL query engine for Hadoop. It is being pushed by MapR, although they are also now supporting Impala. Apache Drill has similar goals to Impala and Presto – fast interactive queries for large datasets, and like these technologies it also requires installation of worker nodes (drillbits). However, unlike Impala and Presto, Drill aims to support multiple backing stores (HDFS, HBase, MongoDB), and has a focus on complex nested datasets (like JSON). Unfortunately, drill is currently only in Alpha, so it is not widely used. Drill Provides:

• ANSI compliant SQL
• Interoperability with several back-end datastores and metadata-stores (Hive, HBase, MongoDB)
• Extension framework for UDFs, storage plugins,

SHOULD YOU USE IT?

Probably not. The project is still in the alpha state, so for now it’s probably not suitable for production use. It’s potentially a project to watch though.

APACHE DRILL VIDEO

 

HAWQ

Now things start to get complicated.

Hawq is a closed-source product from EMC Pivotal offered as part of ‘Pivotal HD’ their proprietary distribution of Hadoop. Pivotal claim that Hawq is the ‘worlds fastest SQL engine on Hadoop’ and that it has been in development for 10 years. However such claims are hard to substantiate. It’s hard to work out exactly what features Hawq provides, but I could glean the following:

• Full SQL syntax support
• Interoperability with Hive and HBase through the Pivotal Xtension Framework(PXF)
• Interoperability with Pivotal’s GemFire XD, their in-memory real-time database backed by HDFS

SHOULD YOU USE IT?

This one is easy. If you use Pivotal’s Hadoop distribution then yes, otherwise no. :-)

HAWQ VIDEO

 

BigSQL by IBM

Big Blue has their own Hadoop Distribution called Big Insights, BigSQL is offered as part of that distribution. BigSQL is used to query data stored in HDFS using both Map-Reduce and something else (unknown) which provides low latency results. From the documents I can access, BigSQL seems to provide:

• JDBC and ODBC drivers
• Broad SQL support
• Presumably a command line client

SHOULD YOU USE IT?

Again, if you’re a customer of IBM then yes you should! Otherwise, probably not. :-)

BIGSQL VIDEO

 

Apache Phoenix

Apache Phoenix is an open-source SQL engine for Apache HBase. The goal of Phoenix is to provide low-latency queries for data stored in HBase via an embeddable JDBC driver. Unlike the other engines we’ve been exploring, Phoenix offers both read and write operations on HBase data. It provides:

• a JDBC driver
• a command-line client
• facilities for bulk-loading data
• the ability to create new tables, or map to existing HBase data

SHOULD YOU USE IT?

If you use HBase there’s no reason not to. Although Hive has the ability to read data from HBase, Phoenix also allows you to write data. It’s unclear whether it is appropriate for production, transactional use, but it’s certainly powerful enough as an analytics tool.

PHOENIX VIDEO

 

Apache Tajo

Apache Tajo is a project to build an advanced data warehousing system on top of HDFS. From my understanding Tajo bills itself as a ‘big data warehouse’, but it seems similar to the other low-latency query engines I’ve already covered. While it has support for external tables and Hive datasets (via HCatalog) it’s focus in on managing the data, providing low-latency access to the data, and providing tools for more traditional ETL. In similar fashion to Impala/Presto/etc it requires that you deploy Tajo-specific worker processes to your datanodes. Tajo provides:

• ANSI SQL compliance
• JDBC Drivers
• Hive metastore integration for access to Hive datasets
• A command line client
• An API for custom functions.

SHOULD YOU USE IT?

While there are some benchmarks that show promising results for Tajo, benchmarks are inherently biased and shouldn’t be fully trusted. The community side of Tajo seems pretty light, and there are no major Hadoop vendors in North America that support it. That said, if you’re in South Korea Gruter is the primary project sponsor and would be a good source of support if you’re on their platform. Overall, if you’re not working with Gruter, it’s a hard sell over more well-known query engines such as Impala or Presto.

TAJO VIDEO

Apache Spark – a Fast Big Data Analytics Engine

Introduction

There are different approaches in big data world to make Hadoop more suitable for ad-hoc, interactive queries and iterative data processing. As it is very well known, Hadoop MapReduce framework is primarily designed for batch processing and that makes it less suitable for ad-hoc data exploration, machine learning processes and the like. Big data vendors are trying to address this challenge by replacing MaReduce with alternatives. In case of SQL on Hadoop, there are various initiatives; Cloudera Impala, Pivotal HAWQ orHortonworks Stinger initiative that aims to improve Hive performance significantly.

Apache Spark is another increasingly popular alternative to replace MapReduce with a more performant execution engine but still use Hadoop HDFS as storage engine for large data sets.

Spark Architecture

From architecture perspective Apache Spark is based on two key concepts; Resilient Distributed Datasets (RDD) and directed acyclic graph (DAG) execution engine. With regards to datasets, Spark supports two types of RDDs: parallelized collections that are based on existing Scala collections and Hadoop datasets that are created from the files stored on HDFS. RDDs support two kinds of operations: transformations and actions. Transformations create new datasets from the input (e.g. map or filter operations are transformations), whereas actions return a value after executing calculations on the dataset (e.g. reduce or count operations are actions).
The DAG engine helps to eliminate the MapReduce multi-stage execution model and offers significant performance improvements.

Figure 1: Spark Architecture

Installing Spark

Spark is written in Scala so before you install Spark, you need to install Scala. Scala binaries can be downloaded from http://www.scala-lang.org.

$ wget http://www.scala-lang.org/files/archive/scala-2.10.4.tgz
$ tar xvf scala-2.10.4.tgz 
$ ln -s scala-2.10.4 scala
$ vi .bashrc
export SCALA_HOME=/home/istvan/scala
export PATH=$SCALA_HOME/bin:$PATH

You can validate your Scala installation by running Scala REPL (Scala command line interpreter), below is an example how to execute the classic HelloWorld program from Scala:

$ scala
Welcome to Scala version 2.10.4 (Java HotSpot(TM) 64-Bit Server VM, Java 1.6.0_32).
Type in expressions to have them evaluated.
Type :help for more information.

scala> object HelloWorld {
     |     def main(args: Array[String]) {
     |         println("Hello, world!")
     |     }
     | }
defined module HelloWorld

scala> HelloWorld.main(null)
Hello, world!

Then you can download Spark binaries from http://spark.apache.org/downloads.html. There are a couple of pre-compiled versions depending on your Hadoop distribution; we are going to use Spark binaries built for Cloudera CDH4 distribution.

$ wget http://d3kbcqa49mib13.cloudfront.net/spark-0.9.0-incubating-bin-cdh4.tgz
$ tar xvf spark-0.9.0-incubating-bin-cdh4.tgz 
$ ln -s spark-0.9.0-incubating-bin-cdh4 spark

Using Spark shell

Now we are ready to run Spark shell which is a command line interpreter for Spark:

$ bin/spark-shell
14/03/31 15:54:51 INFO HttpServer: Using Spark's default log4j profile: org/apache/spark/log4j-defaults.properties
14/03/31 15:54:51 INFO HttpServer: Starting HTTP Server
Welcome to
      ____              __
     / __/__  ___ _____/ /__
    _\ \/ _ \/ _ `/ __/  '_/
   /___/ .__/\_,_/_/ /_/\_\   version 0.9.0
      /_/

Using Scala version 2.10.3 (Java HotSpot(TM) 64-Bit Server VM, Java 1.6.0_32)
Type in expressions to have them evaluated.
Type :help for more information.
....
14/03/31 15:54:59 INFO HttpServer: Starting HTTP Server
...
Created spark context..
Spark context available as sc.

scala>

In our example we are going to process Apache weblogs that support the common logfile format having the following fields: hostname, timestamp, request, HTTP status code and number of bytes. The test file that we are using in this example is based on the public NASA weblog from 1995 August, see http://ita.ee.lbl.gov/html/contrib/NASA-HTTP.html. The file was cleaned and modified to support tab separated format.

Let us assume that you want to count how many hits the NASA web server got in August, 1995. In order to get the result, you can run the following commands in spark-shell:

Spark context available as sc.

scala> val accessLog = sc.textFile("hdfs://localhost/user/cloudera/weblog/NASA_access_log_Aug95")
14/03/31 16:18:16 INFO MemoryStore: ensureFreeSpace(82970) called with curMem=0, maxMem=311387750
14/03/31 16:18:16 INFO MemoryStore: Block broadcast_0 stored as values to memory (estimated size 81.0 KB, free 296.9 MB)
accessLog: org.apache.spark.rdd.RDD[String] = MappedRDD[1] at textFile at :12

scala> accessLog.count()
14/03/31 16:18:24 INFO FileInputFormat: Total input paths to process : 1
14/03/31 16:18:24 INFO SparkContext: Starting job: count at :15
14/03/31 16:18:24 INFO DAGScheduler: Got job 0 (count at :15) with 2 output partitions (allowLocal=false)
...
14/03/31 16:18:26 INFO HadoopRDD: Input split: hdfs://localhost/user/cloudera/weblog/NASA_access_log_Aug95:134217728+27316431
..
14/03/31 16:18:26 INFO SparkContext: Job finished: count at :15, took 2.297566932 s
res0: Long = 1569898

The first command creates an RDD from the NASA Apache access log stored on HDFS (hdfs://localhost/user/cloudera/weblog/NASA_access_log_Aug95). The second command (an action executed on the accessLog RDD) will count the number of lines in the file. Note that the execution time is around 2 seconds.

If you are interested to know how many requests were initiated from one particular server (e.g. beta.xerox.com in our example), you need to execute a filter operation and then you run count action on the filtered dataset, see:

scala> val filteredLog = accessLog.filter(line => line.contains("beta.xerox.com"))

filteredLog: org.apache.spark.rdd.RDD[String] = FilteredRDD[2] at filter at :14

scala> filteredLog.count()
14/03/31 16:19:35 INFO SparkContext: Starting job: count at :17
14/03/31 16:19:35 INFO DAGScheduler: Got job 1 (count at :17) with 2 output partitions (allowLocal=false)
...
14/03/31 16:19:35 INFO Executor: Running task ID 2
...
14/03/31 16:19:35 INFO HadoopRDD: Input split: hdfs://localhost/user/cloudera/weblog/NASA_access_log_Aug95:0+134217728
14/03/31 16:19:36 INFO Executor: Serialized size of result for 2 is 563
1...
14/03/31 16:19:36 INFO HadoopRDD: Input split: hdfs://localhost/user/cloudera/weblog/NASA_access_log_Aug95:134217728+27316431
...
14/03/31 16:19:37 INFO DAGScheduler: Stage 1 (count at :17) finished in 1.542 s
14/03/31 16:19:37 INFO SparkContext: Job finished: count at :17, took 1.549706892 s
res1: Long = 2318

You can also execute more complex logic and define your own functions, thanks to Scala language capabilities. Let us assume that we need to calculate the total number of bytes generated by the given NASA web server in August, 1995. The number of bytes is the last (5th) field in the lines in the weblog file and unfortunately, there are cases when the field is ‘-‘, not an integer value as it would be expected. The standard toInt Scala String function throws an exception if you want to convert a non-numeric value into Integer. Thus we need to be able to identify whether a given string is a number or not and if not, we need to return 0. This requires a custom function (convertToInt) that will extend the standard String Scala class and will be made available for the String data type. Then we can use this custom function and the Spark standard RDD operations to calculate the total number of  bytes generated by the NASA webserver.

scala> def isNumeric(input: String): Boolean = input.forall(_.isDigit)
isNumeric: (input: String)Boolean


scala> class StringHelper(s:String) {
     |    def convertToInt():Int = if (isNumeric(s)) s.toInt else 0
     | }
defined class StringHelper


scala> implicit def stringWrapper(str: String) = new StringHelper(str)
warning: there were 1 feature warning(s); re-run with -feature for details
stringWrapper: (str: String)StringHelper


scala> "123".convertToInt
res2: Int = 123

scala> "-".convertToInt
res4: Int = 0

scala> accessLog.map(line=>line.split("\t")).map(line=>line(4).convertToInt).sum39 INFO SparkContext: Starting job: sum at :21
14/04/01 13:12:39 INFO DAGScheduler: Got job 2 (sum at :21) with 2 output partitions (allowLocal=false)
...
14/04/01 13:12:40 INFO HadoopRDD: Input split: hdfs://localhost/user/cloudera/weblog/NASA_access_log_Aug95:0+134217728
... 
14/04/01 13:12:45 INFO SparkContext: Job finished: sum at :21, took 5.53186208 s
res4: Double = 2.6828341424E10

As you can see, the result is 26,8GB and the calculation was executed in 5.5 seconds.

Programming Spark

In addition to spark-shell that can be used to execute operations interactively, you can also write and build your code using Scala, Java or Python programming languages. Let us take an example how you can implement your weblog application in Scala.

In order to build your application, you need to follow the directory structure as shown below:

./
./simple.sbt
./src/main/scala/WeblogApp.scala
./project/build.properties
./sbt

You can copy sbt build tool from your Spark home directory (cp -af $SPARK_HOME/sbt/* ./sbt).

The simple.sbt build file should look something like this:

name := "Weblog Project"

version := "1.0"

scalaVersion := "2.10.4"

libraryDependencies += "org.apache.spark" %% "spark-core" % "0.9.0-incubating"

resolvers += "Akka Repository" at "http://repo.akka.io/releases/"

libraryDependencies += "org.apache.hadoop" % "hadoop-client" % "2.0.0-cdh4.4.0"

And the WeblogApp.scala code is as follows:

import org.apache.spark._

object WeblogApp {
    def main(args: Array[String]) {
        val file = "hdfs://localhost/user/cloudera/weblog/NASA_access_log_Aug95";
        val sc = new SparkContext("local", "WeblogApp",
                                  System.getenv("SPARK_HOME"), 
                                  SparkContext.jarOfClass(this.getClass))
        val accessLog = sc.textFile(file)

        println("Number of entries: " + accessLog.count())
    }
}

Then you can build and run the application using the Scala sbt build tool:

$ sbt/sbt package
$ sbt/sbt run
Launching sbt from sbt/sbt-launch-0.12.4.jar
[info] Loading project definition from /home/istvan/project
[info] Set current project to Weblog Project (in build file:/home/istvan/)
[info] Running WeblogApp 
...
14/04/01 14:48:58 INFO SparkContext: Starting job: count at WeblogApp.scala:12
14/04/01 14:48:58 INFO DAGScheduler: Got job 0 (count at WeblogApp.scala:12) with 2 output partitions (allowLocal=false)
...
14/04/01 14:49:01 INFO SparkContext: Job finished: count at WeblogApp.scala:12, took 2.67797083 s
Number of entries: 1569898
14/04/01 14:49:01 INFO ConnectionManager: Selector thread was interrupted!
[success] Total time: 9 s, completed 01-Apr-2014 14:49:01

Conclusion

Apache Spark has started gaining significant momentum and considered to be a promising alternative to support ad-hoc queries and iterative processing logic by replacing MapReduce. It offers interactive code execution using Python and Scala REPL but you can also write and compile your application in Scala and Java. There are various tools running on top of Spark such as Shark (SQL on Hadoop), MLib (machine learning), Spark Streaming and GraphX.It will be interesting to see how it evolves.