Tag: Spark SQL

The Year in SQL Engines

As an addendum to my year-end review of machine learning and deep learning, I offer this survey of SQL engines. SQL is the most widely used language for data science according to O’Reilly’s 2016 Data Science Salary Survey. Most projects require at least some SQL operations, and many need nothing but SQL.

This review covers six open source leaders: Hive, Impala, Spark SQL, Drill, HAWQ, and Presto; plus, for completeness, Calcite, Kylin, Phoenix, Tajo, and Trafodion. Omitted: two commercial options, Oracle Big Data SQL and IBM Big SQL, which IBM has not yet rebranded as “Watson SQL.”

(A reader asks: What about Druid? My response: erm. On inspection, I agree that Druid belongs in this category, so check it out.)

I use the term ‘SQL Engine’ loosely. Hive, for example, is not an engine; it’s a framework that uses the MapReduce, Tez, or Spark engines to run queries. And it doesn’t run SQL; it runs HiveQL, an SQL-like language that closely approximates SQL. ‘SQL-in-Hadoop’ is also inapt; while Hive and Impala work primarily with Hadoop, Spark, Drill, HAWQ, and Presto also work with a wide variety of other data storage systems.

Unlike relational databases, SQL engines operate independently of the data storage system. In contrast, relational databases bundle the query engine and storage into a single tightly coupled system, which permits certain types of optimization. Uncoupling them, on the other hand, provides greater flexibility, though at the potential loss of performance.

Figure 1, below, shows the relative popularity of the leading SQL engines according to DB-Engines, a website maintained by the Austrian consultancy Solid IT. DB-engines computes a monthly popularity score for more than 200 database systems. The score reflects search engine queries; mentions in online discussions; job offers; mentions in professional profiles, and tweets.

Figure 1

screen-shot-2017-01-31-at-1-04-43-pm
Source: DB-Engines, January 2017 http://db-engines.com/en/ranking

Although Impala, Spark SQL, Drill, Hawq, and Presto consistently beat Hive on measures such as runtime performance, concurrency, and throughput, Hive remains the most popular (at least by the DB-Engines metric). There are three reasons why that is so:

— Hive is the default option for SQL in Hadoop, supported in every distribution. The others align with specific vendors and cater to niche users.

— Hive has closed the performance gap to the other engines. Most of the Hive alternatives launched in 2012 when analysts would rather kill themselves than wait for a Hive query to finish. But while Impala, Spark, Drill, et.al. ran away like rabbits back then, Hive just kept chugging along, tortoise-like, with incremental improvements. Today, while Hive is not the fastest choice, it’s a lot better than it was five years ago.

— While bleeding-edge speed is cool, most organizations know that the world does not end if a junior marketing manager has to wait ten seconds to find out if the chicken wings outperformed the buffalo burgers in the Duxbury restaurant last Tuesday.

As you can see in Figure 2, below, the top SQL engines compete well for user interest compared to leading commercial data warehouse appliances.

Figure 2

screen-shot-2017-01-31-at-2-27-15-pm
Source: DB-Engines, January 2017 http://db-engines.com/en/ranking

The best measure of health for an open source project is the size of its active developer community. Hive and Presto have the largest base of contributors, as shown in Figure 3, below. (Data for Spark SQL is unavailable.)

Figure 3

screen-shot-2017-01-31-at-2-52-27-pm
Source: Open Hub https://www.openhub.net/

In 2016, Cloudera, Hortonworks, Kognitio, and Teradata waded into the Battle of the Benchmarks Tony Baer summarizes. I’m sure that you will be shocked to learn that the vendor’s preferred SQL engine outperformed the others in each of these studies, which begs the question: are benchmarks bullshit?

AtScale‘s biannual benchmark is not BS. AtScale, a BI startup, markets software that brokers between BI front ends and SQL backends. The company’s software is engine-neutral — it seeks to run on as many as possible — and its broad experience in BI gives the testing a real-world flavor.

AtScale’s key findings from its most recent round, which included Hive, Impala, Spark SQL, and Presto:

— All four engines successfully ran AtScale’s BI benchmark queries.

— Each engine has its own performance “sweet spot” depending on data volume, query complexity, and concurrent users.

– Impala and Spark SQL outperform the others in queries against small data sets

– On large data sets, Impala and Spark SQL handle complex joins better than the others

– Impala and Presto demonstrate the best results in concurrency tests

— All engines showed 2X-4X performance gains in the six months since AtScale’s previous benchmark.

Alex Woodie reports on the test results; Andrew Oliver analyzes.

Let’s dive into the individual projects.

Apache Hive

Apache Hive was the first SQL framework in the Hadoop ecosystem. Engineers at Facebook introduced Hive in 2007 and donated the code to the Apache Software Foundation in 2008; in September 2010, Hive graduated to top-level Apache project status. Every major player in the Hadoop ecosystem distributes and supports Hive, including Cloudera, MapR, Hortonworks, and IBM. Amazon Web Services offers a modified version of Hive as a cloud service in Elastic MapReduce (EMR).

Early releases of Hive used MapReduce to run queries. Complex queries required multiple passes through the data, which impaired performance. As a result, Hive was not suitable for interactive analysis. Led by Hortonworks, the Stinger initiative markedly enhanced Hive’s performance, notably through the use of Apache Tez, an application framework that delivers streamlined MapReduce code. Tez and ORCfile, a new storage format, produced a significant speedup for Hive queries.

Cloudera Labs spearheaded a parallel project to re-engineer Hive’s back end to run on Apache Spark. After an extended beta, Cloudera released Hive-on-Spark to general availability in early 2016.

More than 100 individuals contributed to Hive in 2016. The team announced Hive 2.0 in February and Hive 2.1 in June. Hive 2.0 includes improvements to several improvements to Hive-on-Spark, plus performance, usability, supportability and stability enhancements. Hive 2.1 includes Hive LLAP (“Live Long and Process”), which combines persistent query servers and optimized in-memory caching for high performance. The team claims a 25X speedup.

In September, the Hivemall project entered the Apache Incubator, as I noted in Part Two of my machine learning year-end roundup. Originally developed by Treasure Data and donated to the Apache Software Foundation, Hivemall is a scalable machine learning library implemented as a collection of Hive UDFs designed to run in Hive, Pig or Spark SQL with MapReduce, Tez or Spark. The team plans an initial release in Q1 2017.

Apache Impala

Cloudera launched Impala, an open source MPP SQL engine, in 2012, as a high-performance alternative to Hive. Impala works with HDFS and HBase, and it leverages Hive metadata; however, it bypasses MapReduce to run queries. Mike Olson, Cloudera’s Chief Strategy Officer,

Mike Olson, Cloudera’s Chief Strategy Officer, argued in late 2013 that Hive’s architecture was fundamentally flawed. In Olson’s view, developers could only deliver high-performance SQL with a whole new approach, exemplified by Impala. In 2014 Cloudera released a series of benchmarks in January, May, and September. In these tests, Impala showed progressive improvement in query runtime, and significantly outperformed Hive on Tez, Spark SQL, and Presto. In addition to running fast, Impala performed particularly well in concurrency, throughput, and scalability.

In 2015, Cloudera donated Impala to the Apache Software Foundation, where it entered the Apache Incubator program. Cloudera, MapR, Oracle and Amazon Web Services distribute Impala;  Cloudera, MapR, and Oracle provide commercial build and installation support.

Impala made steady progress in the Apache Incubator in 2016. The team cleaned up the code, ported it to Apache infrastructure and delivered Release 2.7.0, its first Apache release in October. The new version includes performance and scalability improvements, as well as some other minor enhancements.

In September, Cloudera published results of a study that compared Impala to Amazon Web Services’ Redshift columnar database. The report is interesting reading, though subject to the usual caveats about vendor benchmarks.

Spark SQL

Spark SQL is a Spark component for structured data processing. The Apache Spark team launched Spark SQL in 2014 and absorbed Shark, an early Hive-on-Spark project. It quickly became the most widely used Spark module.

Spark SQL users can run SQL queries, read data from Hive, or use it as means to create Spark Datasets and DataFrames. (Datasets are distributed collections of data; DataFrames are Datasets organized into named columns.) The Spark SQL interface provides Spark with information about the structure of the data and operations to be performed; Spark’s Catalyst optimizer uses this information to construct an efficient query.

In 2015, Spark’s machine learning developers introduced the ML API, a package that leveraged Spark DataFrames instead of the lower-level Spark RDD API. This approach proved to be attractive and fruitful; in 2016, with Release 2.0, the Spark team placed the RDD-based API in maintenance mode. The DataFrames API is now the primary interface for Spark machine learning.

Also in 2016, the team released Structured Streaming, in an Alpha release as of Spark 2.1.0. Structured Streaming is a stream processing engine built on Spark SQL. Users can query streaming data sources in the same manner as static sources, and they can combine streaming and static sources in a single query. Spark SQL runs the query continuously and updates results as streaming data arrives. Structured Streaming delivers exactly-once fault-tolerance guarantees through checkpointing and Write Ahead Logs.

Apache Drill

In 2012, a group led by MapR, one of the leading Hadoop distributors, proposed to build an open-source version of Google’s Dremel, a distributed system for interactive ad-hoc analysis. They named the project Apache Drill. Drill languished in the Apache Incubator for more than two years, finally graduating in late 2014. The team delivered its 1.0 release in 2015.

MapR distributes and supports Apache Drill.

More than 50 individuals contributed to Drill in 2016. The team delivered five dot releases in 2016. Key enhancements include:

  • Web authentication
  • Support for the Apache Kudu columnar database
  • Support for HBase 1.x
  • Dynamic UDF support

Two key Drill contributors left MapR to start Dremio in 2015; the startup remains in stealth mode.

Apache HAWQ

Pivotal Software introduced HAWQ as a commercially licensed high-performance SQL engine in 2012 and attempted to market it with minimal success. Changing strategy, Pivotal donated the project to Apache in June 2015, and it entered the Apache Incubator program in September 2015.

Fifteen months later, HAWQ remains in the Incubator. The team released HAWQ 2.0.0.0 in December, with a load of bug fixes. I suspect the project will graduate in 2017.

One small point in HAWQ’s favor is its support for Apache MADlib, the machine-learning-in-SQL project that is also still in the Incubator. The combination of HAWQ and MADlib should be a nice consolation to the folks who bought Greenplum and wonder what the hell happened.

Presto

Facebook engineers initiated the Presto project in 2012 as a fast interactive alternative to Hive. Rolled out in 2013, the software successfully supported more than a thousand Facebook users and more than 30,000 queries per day on petabytes of data. Facebook released Presto to open source in 2013.

Presto supports ANSI SQL queries across a range of data sources, including Hive, Cassandra, relational databases or proprietary file systems (such as Amazon Web Services’ S3.)  Presto queries can federate data from multiple sources.  Users can submit queries from C, Java, Node.js, PHP, Python, R and Ruby.

Airpal, a web-based query tool developed by Airbnb, offers users the ability to submit queries to Presto through a browser. Qubole provides a managed service for Presto. AWS delivers a Presto service on EMR.

In June 2015, Teradata announced plans to develop and support the project.  Under an announced three-phase program, Teradata proposed to integrate Presto into the Hadoop ecosystem, enable operation under YARN and enhance connectivity through ODBC and JDBC. Teradata offers its own distribution of Presto, complete with a data sheet. In June, Teradata announced the certification of Information Builders, Looker, Qlik, Tableau, and ZoomData, with MicroStrategy and Microsoft Power BI on the way.

Presto is a very active project, with a vast and vibrant contributor community. The team cranks out releases faster than Miki Sudo eats hot dogs — I count 42 releases in 2016. Teradata hasn’t bothered to summarize what’s new, and I don’t plan to sift through 42 sets of release notes, so let’s just say it’s better.

Other Apache Projects

There are five other SQL-ish projects in the Apache ecosystem.

Apache Calcite

Apache Calcite is an open source framework for building databases. It includes:

— A SQL parser, validator and JDBC driver

— Query optimization tools, including a relational algebra API, rule-based planner, and a cost-based query optimizer.

Apache Hive uses Calcite for cost-based query optimization, while Apache Drill and Apache Kylin use the SQL parser.

The Calcite team pushed out five releases in 2016, with bug fixes and new adapters for Cassandra, Druid, and Elasticsearch.

Apache Kylin

Apache Kylin is an OLAP engine with a SQL interface. Developed by eBay and donated to Apache, Kylin graduated to top-level status in 2015.

A startup named Kyligence launched in 2016; it offers commercial support and a data warehousing product called KAP, FWIW. While the company has no funding listed in Crunchbase, a source tells me that it has strong backing and a large office in Shanghai.

Apache Phoenix

Apache Phoenix is a SQL framework that runs on HBase and bypasses MapReduce. Salesforce developed the software and donated it to Apache in 2013. The project graduated to top-level status in May 2014. Hortonworks includes Phoenix in the Hortonworks Data Platform. Since the leading SQL engines all work with HBase, it’s not clear why we need Phoenix.

Apache Tajo

Apache Tajo is a fast SQL data warehousing framework introduced in 2011 by Gruter, a Big Data infrastructure company, and donated to Apache in 2013. Tajo graduated to top level status in 2014. The project has attracted little interest from prospective users and contributors outside of Gruter’s primary market in South Korea. Other than a brief mention by Gartner’s Nick Heudecker, the project isn’t on anyone’s dashboard.

Apache Trafodion

Apache Trafodion is another SQL-on-HBase project, conceived by HP Labs, which tells you pretty much all you need to know. HP launched Trafodion in June 2014, a month after Apache Phoenix graduated to production. Six months later, it dawned on HP executives that there might be limited commercial potential for another SQL-on-HBase engine — I can see the facepalms — so they donated the project to Apache, where it entered the Incubator in May 2015.

Trafodion promises to be a transactional database if it ever gets out of incubation. Unfortunately, there are lots of options in that space, and the only competitive benefit the development team can articulate seems to be “it’s open source, so it’s cheap.”

Big Analytics Roundup (February 29, 2016)

Happy Leap Day.  Tachyon’s rebranding as Alluxio, release of CaffeOnSpark and GA for Google Cloud Dataproc lead the hard news this week.  The Alluxio announcement has inspired big thinkers to share big thoughts.  And, we have a nice crop of explainers.  Scroll down to the bottom for another SQL on Hadoop benchmark.

Explainers

— In SearchDataManagement, Jack Vaughn explains Spark 2.0.

— In Datanami, Alex Woodie explains Structured Streaming in Spark 2.0.

— MapR’s Jim Scott explains Spark accumulators.   Jim also explains Spark Streaming.

— DataArtisans’ Fabian Hueske introduces Flink.

— In SlideShare, Julian Hyde explains streaming SQL.

— Wes McKinney explains why pandas users should be excited about Apache Arrow.

— On her blog, Paige Roberts explains Project Tungsten, complete with pictures.

— Someone from Dremio explains Drillix, which is what you get when you combine Apache Phoenix and Apache Drill. (h/t Hadoop Weekly).

Perspectives

— In TheNextPlatform, Timothy Prickett Morgan argues that Tachyon Caching (Alluxio) is bigger than Spark

— In SiliconAngle, Maria Deutscher opines that Alluxio (née Tachyon) could replace HDFS for Spark users.

— In The New Stack, Susan Hall speculates that Apache Arrow’s columnar data layer could accelerate Spark and Hadoop.  She means Hadoop in a general way, e.g. the Hadoop ecosystem.

— On the Dataiku blog, “Caroline” interviews John Kelly, Managing Director of Berkeley Research Group and asks him questions about data science.  Left unanswered: is it “Data-ikoo” or “Day-tie-koo?”

— Alpine Data Labs’ Steven Hillion ruminates on success.  He’d be better off ruminating on “how to raise your next round of venture capital.”

— Max Slater-Robins opines that Microsoft is inventing the future, which is even better than winning the internet.

— In ZDNet, Andrew Brust wonders if Databricks is vying for a full analytics stack, citing the new Dashboard feature as cause for wonder.  He’s just trolling.

— In Search Cloud Applications, Joel Shore opines that streaming analytics is replacing complex event processing, which makes sense.   He further opines that Flink will displace Spark for streaming, which doesn’t make sense.   Shore interviews IBM’s Nagui Halim about streaming here.

Open Source Announcements

— Alluxio (née Tachyon) announces Release 1.0.0.  Alluxio is open source software distributed through Git under an Apache license, but is not an Apache project.  Yet.  Release 1.0 includes frameworks for MapReduce, Spark, Flink and Zeppelin.  Daniel Gutierrez reports.

— Yahoo releases CaffeOnSpark, a distributed deep learning package.  Caffe is one of the better-known deep learning packages, with a track record in image recognition.  Software is available on Git.  For more information, see the Wiki.  Alex Handy reports; Charlie Osborne reports.

— RapidMiner China announces availability of an extension for deep learning engine DL4J.  The extension is open source, and works with the open source version of RapidMiner.  DL4J sponsor Skymind collaborated.

Commercial Announcements

–Tachyon Nexus, the commercial venture founded to support Tachyon, the memory-centric virtual distributed storage system, announces that it has rebranded as Alluxio.

— Google announces general availability for its Cloud Dataproc managed service for Spark and Hadoop.

Funding Announcements

Health analytics vendor Health Catalyst lands a $70M Series E round.

AtScale Benchmarks SQL-on-Hadoop Engines

On the AtScale blog, Trystan Leftwich summarizes results from a benchmark test of Hive on Tez (1.2/0.7), Cloudera Apache Impala (2.3) and Spark SQL (1.6).  The AtScale team tested Impala and Spark with Parquet and Hive on Tez with ORC.  For test cases, the team used TPC-H data arranged in a star schema, and ran 13 queries in each SQL engine multiple times, averaging the results.

While Hortonworks recommends ORC with Hive/Tez, there are published cases where users achieved good results with Hive/Tez on Parquet.  Since the storage format has a big impact on SQL performance, I would have tested Hive/Tez on Parquet as well.  AtScale did not respond to queries on this point.

Key findings:

  • All three engines performed about the same on single-table queries, and on queries joining three small tables.
  • Spark and Impala ran faster than Hive on queries joining three large tables.
  • Spark ran faster than Impala on queries joining four or more tables.

The team ran the same tests with AtScale’s commercial caching technology, with significant performance improvements for all three engines.

In concurrency testing, Impala performed much better than Hive or Spark.

Details of the test available in a white paper here (registration required).

Looking Ahead: Big Analytics in 2016

Every year around this time I review last year’s forecast and publish some thoughts about the coming year.

2015 Assessment

First, a brief review of my predictions for 2015:

(1) Apache Spark usage will explode.

Nailed it.

(2) Analytics in the cloud will take off.

In 2015, all of the leading cloud platforms — AWS, Azure, IBM and Google — released new tools for advanced analytics and machine learning.  New cloud-based providers specializing in advanced analytics, such as Qubole and Domino Data, emerged.

Cloud platform providers do not break out revenue by workload, so it’s difficult to measure analytics activity in the cloud; anecdotally, though, there are a growing number of analysts, vendors and service providers whose sole platform is the cloud.

(3) Python will continue to gain on R as the preferred open source analytics platform.

While Python continues to add functionality and gain users, so does R, so it’s hard to say that one is gaining on the other.

(4) H2O will continue to win respect and customers in the Big Analytics market.

In 2015, H2O doubled its user base, expanded its paid subscriber base fourfold and landed a $20 million “B” round.  Not bad for a company that operates on a true open source business model.

(5) SAS customers will continue to seek alternatives.

Among analytic service providers (ASPs) the exit from SAS is a stampede.

With a half dozen dot releases, SAS’ distributed in-memory products are stable enough that they are no longer the butt of jokes.  Customer adoption remains thin; customers are loyal to SAS’ legacy software, but skeptical about the new stuff.

2016 Themes

Looking ahead, here is what I see:

(1) Spark continues its long march into the enterprise.

With Cloudera 6, Spark will be the default processing option for Cloudera workloads.  This does not mean, as some suggest, that MapReduce is dead; it does mean that a larger share of new workloads will run on Spark.  Many existing jobs will continue to run in MapReduce, which works reasonably well for embarrassingly parallel workloads.

Hortonworks and MapR haven’t followed Cloudera with similar announcements yet, but will do so in 2016.  Hortonworks will continue to fiddle around with Hive on Tez, but will eventually give up and embrace Hive on Spark.

SAS will hold its nose and support Spark in 2016.  Spark competes with SAS’ proprietary back end, but it will be forced to support Spark due to its partnerships with the Hadoop distributors.  Analytic applications like Datameer and Microsoft/Revolution Analytics ScaleR that integrate with Hadoop through MapReduce will rebuild their software to interface with Spark.

Spark Core and Spark SQL will remain the most widely used Spark components, with general applicability across many use cases.  Spark MLLib suffers from comparison with alternatives like H2O and XGBoost; performance and accuracy need to improve.  Spark Streaming faces competition from Storm and Flink; while the benefits of “pure” streaming versus micro-batching are largely theoretical, it’s a serious difference that shows up in benchmarks like this.

With no enhancements in 2015, Spark GraphX is effectively dead.  The project leadership team must either find someone interested in contributing, fold the library into MLLib, or kill it.

(2) Open source continues to eat the analytics software world.

If all you read is Gartner and Forrester, you may be inclined to think that open source is just a blip in the market.  Gartner and Forrester ignore open source analytics for two reasons: (1) they get paid by commercial vendors, and (2) users don’t need “analysts” to tell them how to evaluate open source software.  You just download it and check it out.

Surveys of actual users paint a different picture.  Among new grads entering the analytics workforce, using open source is as natural as using mobile phones and Yik Yak; big SAS shops have to pay to send the kids to training.  The best and brightest analysts use open source tools, as shown by the 2015 O’Reilly Data Science Salary Survey;  while SAS users are among the lowest paid analysts, they take consolation from knowing that SPSS users get paid even less.

IBM’s decision in 2015 to get behind Spark exemplifies the movement towards open source.  IBM ranks #2 behind SAS in advanced analytics software revenue, but chose to disrupt itself by endorsing Spark and open-sourcing SystemML.  IBM figures to gain more in cloud and services revenue than it loses in cannibalized software sales.  It remains to be seen how well that will work, but IBM knows how to spot a trend when it sees it.

Microsoft’s acquisition of Revolution Analytics in 2015 gives R the stamp of approval from a company that markets the most widely implemented database (SQL Server) and the most widely used BI tool (Excel).  As Microsoft rolls out its R server and SQL-embedded R, look for a big jump in enterprise adoption.  It’s no longer possible for folks to dismiss R as some quirky tool used by academics and hobos.

The open source business model is also attracting capital.  Two analytics vendors with open source models (H2O and RapidMiner) recently landed funding rounds, while commercial vendors Skytree and Alpine languish in the funding doldrums and cut headcount.  Palantir and Opera, the biggest dogs in the analytics startup world, also leverage open source.

Increasingly, the scale-out distributed back end for Big Analytics is an open source platform, where proprietary architecture sticks out like a pimple.  Commercial software vendors can and will thrive when they focus on the end user.  This approach works well for AtScale, Alteryx, RapidMiner and ZoomData, among others.

(3) Cloud emerges as the primary platform for advanced analytics.

By “cloud” I mean all types of cloud: public, private, virtual private and hybrid, as well as data center virtualization tools, such as Apache Mesos.  In other words, self-service elastic provisioning.

High-value advanced analytics is inherently project-oriented and ad-hoc; the most important questions are answered only once.  This makes workloads for advanced analytics inherently volatile.  They are also time-sensitive and may require massive computing resources.

This combination  — immediate need for large-scale computing resources for a finite period — is inherently best served by some form of cloud.  The form of cloud an organization chooses will depend on a number of factors, such as where the source data resides, security concerns and the organization’s skills in virtualization and data center management.  But make no mistake: organizations that do not leverage cloud computing for advanced analytics will fall behind.

Concerns about cloud security for advanced analytics are largely bogus: rent-seeking apologetics from IT personnel who (rightly) view the cloud as a threat to their fiefdom.  Sorry guys — the biggest data breaches in the past two years were from on-premises systems.  Arguably, data is more secure in one of the leading clouds than it is in on premises.

For more on this, read my book later this year. 🙂

(4) Automated machine learning tools become mainstream.

As I’ve written elsewhere, automated machine learning is not a new thing.  Commercial and open source tools that automate modeling in various ways have been available since the 1980s.  Most, however, automated machine learning by simplifying the problem in ways that adversely impact model quality.  In 2016, software will be available to enterprises that delivers expert-level predictive models that win Kaggle competitions.

Since analysts spend 80% of their time data wrangling, automated machine learning tools will not eliminate the hiring crunch in advanced analytics; one should be skeptical of vendor claims that “it’s so easy that even a caveman can do it.”  The primary benefit of automation will be better predictive models built consistently to best practices.  Automation will also expand the potential pool of users from hardcore data scientists to “near-experts”, people with business experience or statistical training who are not skilled in programming languages.

(5) Teradata continues to struggle.

Listening to Teradata’s Q3 earnings call back in November, I thought of this:

100_anniversary_titanic_sinking_by_esai8mellows-d4xbme8

CEO Mike Koehler, wiping pie from his face after another quarterly earnings fail, struggled to explain a coherent growth strategy.  It included (a) consulting services; (b) Teradata software on AWS; (c) Aster on commodity hardware.

Well, that dog won’t hunt.

— Teradata’s product sales drive its consulting revenue.  No product sales, no consulting revenue.   Nobody will ever hire Teradata for platform-neutral enterprise Big Data consulting projects, so without a strategy to build product sales, consulting  revenue won’t grow either.

— Teradata’s principal value added is its ability to converge software and hardware into an integrated appliance.  By itself, Teradata software itself is nothing special; there are plenty of open source alternatives, like Apache Greenplum.  Customers who choose to build a data warehouse on AWS have many options, and Teradata won’t be the first choice.  Meanwhile, IBM, Microsoft and Oracle are light years ahead of Teradata delivering true hybrid cloud databases.

— Aster on commodity hardware is a SQL engine with some prebuilt apps.  It runs through MapReduce, which was kind of cool in 2012 but DOA in today’s market: customers who want a SQL engine that runs on commodity hardware have multiple open source options, including Presto, which Teradata also embraces.

Meanwhile, Teradata’s leadership team actually spent time with analysts talking about the R&D tax credit, which seemed like shuffling deck chairs.  The stock is worth about a third of its value in 2012 because the company has repeatedly missed earnings forecasts, and investors have no confidence in current leadership.

At current market value, Teradata is acquisition bait, but it’s not clear who would buy it.  My money’s on private equity, who will cut headcount by half and milk the existing customer base.   There are good people at Teradata; I would advise them all to polish their resumes.

2015 in Big Analytics

Looking back at 2015, a few stories stand out:

  • Steady progress for Spark, punctuated by two big announcements.
  • Solid growth in cloud-based machine learning, led by Microsoft.
  • Expanding options for SQL and OLAP on Hadoop.

In 2015, the most widely read post on this blog was Spark is Too Big to Fail, published in April.  I wrote this post in response to a growing chorus of snark about Spark written by folks who seemed to know little about the project and its goals.

IBM Embraces Spark

IBM’s commitment to Spark, announced on Jun 15, lit up the crowds gathered in San Francisco for the Spark Summit.  IBM brings a number of things to Spark: deep pockets to build a community, extensive technical resources and a large customer base.  It also brings a clutter of aging and partially integrated products, an army of suits and no less than 164 Vice Presidents whose titles include the words “Big Data.”

When IBM announced its Spark initiative I joked that somewhere in the bowels of IBM, someone will want to put Spark on a mainframe.  Color me prophetic.

It’s too early to tell what substantive contributions IBM will make to Spark.  Unlike Mesosphere, Typesafe, Tencent, Palantir, Cloudera, Hortonworks, Huawei, Shopify, Netflix, Intel, Yahoo, Kixer, UC Berkeley and Databricks, IBM did not help test Release 1.5 in September.  This is a clear miss, given the scope of IBM’s resources and the volume of hype it puts out about its commitment to the project.

All that said, IBM brings respectability, and the assurance that Spark is ready for prime time.  This is priceless.  Since IBM’s announcement, we haven’t heard a peep from the folks who were snarking at Spark earlier this year.

Cloudera Announces “One Platform” Initiative

In September, Cloudera announced its One Platform initiative to unify Spark and Hadoop, an announcement that surprised everyone who thought Spark and Hadoop were already pretty well integrated.  As with the IBM announcement, the symbolism matters.  Some analysts took this announcement to mean that Cloudera is replacing MapReduce with Spark, which isn’t exactly true.  It’s fairer to say that in Cloudera’s vision, Hadoop users will rely more on Spark in the future than they do today, but MapReduce is not dead.

The “One Platform” positioning has more to do with Cloudera moving to stem the tide of folks who use Spark outside of Hadoop.  According to Databricks’ recent Spark user survey, only 40% use Spark under YARN, with the rest running in a freestanding cluster or on Mesos.  It’s an understandable concern for Cloudera; I’ve never heard a fish seller suggest that we should eat less fish.  But if Cloudera thinks “One Platform” will stem that tide, it is mistaken.  It all boils down to use cases, and there are many use cases for Spark that don’t need Hadoop’s baggage.

Microsoft Builds Credibility in Analytics

In 2015, Microsoft took some big steps to demonstrate that it offers serious solutions for analytics.  The acquisition of Revolution Analytics, announced in January, was the first step; in one move, Microsoft acquired a highly skilled team and valuable software assets.  Since the acquisition, Microsoft has rolled Revolution’s enhanced R distribution into SQL Server and Azure, opening both platforms to the large and growing R community.

Microsoft’s other big move, in February, was the official launch of Azure Machine Learning (AML).   First released in beta in June 2014, AML is both easy to use and powerful.  The UI is simple to understand, and documentation is excellent; built-in analytic functionality is very rich, and the tool is extensible with custom R or Python scripts.  Microsoft’s trial user program is generous, and clearly designed to encourage adoption and use.

Azure Machine Learning contrasts markedly with Amazon Machine Learning.  Amazon’s offering remains a skeleton, with minimal functionality and an API only a developer could love.  Microsoft is clearly making a play for the data science market as a way to leapfrog Amazon.  If analytic capabilities are driving your choice of cloud platform, Azure is by far your best option.

SQL Engines Proliferate

At the beginning of 2015, there were two main options for SQL on Hadoop: Hive for batch SQL and Impala for interactive SQL.  Spark SQL was still in Alpha; Drill was a curiosity; and Presto was something used at Facebook.

Several things happened during the year:

  • Hive on Tez established rough performance parity with the fast SQL engines.
  • Spark SQL went to general release, stabilized, and rolled out the DataFrames API.
  • MapR promoted Drill, and invested in improvements to the software.  Also, MapR’s Drill team spun off and started Dremio to provide commercial support.
  • Cloudera donated Impala to open source, and Pivotal donated Hawq.
  • Teradata placed its chips on Presto.

While it’s great to see so many options emerge, Hive continues to win actual evaluations.  Given Hive’s large user and contributor base and existing stock of programs, it’s unclear how much traction Hive alternatives have now that Hive on Tez offers competitive performance.  Obviously, Cloudera doesn’t think Impala offers a competitive advantage anymore, or they would not have donated the assets to Apache.

The other big news in SQL is TPC’s release of a benchmarking standard for decision support with Big Data.

OLAP on Hadoop Gets Real

For folks seeking to perform dimensional analysis in Hadoop, 2015 delivered not one but two options.  The open source option, Apache Kylin, originally an eBay project, just recently graduated to Apache top level status.  Adoption is limited at present, but any project used by eBay and Baidu is worth a look.

The commercial option is AtScale, a company that emerged from stealth in April.  Unlike BI-on-Hadoop vendors like Datameer and Pentaho, AtScale provides a dimensional layer designed to work with existing BI tools.  It’s a nice value proposition for companies that have already invested big time in BI tools, and don’t want to add another UI to the mix.

Funding for Machine Learning

H2O.ai’s recently announced B round is significant for a couple of reasons.  First, it validates H2O.ai’s true open source business model; second, it confirms the continued growth and expansion of the user base for H2O as well as H2O.ai’s paid subscription base.

Like Sherlock Holmes’ dog that did not bark, two companies are significant because they did not procure funding in 2015:

  • Skytree, whose last funding round closed in April 2013, churned its executive team and rebranded a couple of times.  It finally listed some new customers; interestingly, some are investors and others are affiliated with members of Skytree’s Board.
  • Alpine Data Labs, last funded in November 2013, struggled to distance itself from the Pivotal ecosystem.  Designed to run on Greenplum, Alpine offers limited functionality on Hadoop, which makes it unclear how this company survives.

Palantir continued to suck up capital like a whale feeding on krill.

Google TensorFlow

Google open sourced TensorFlow, so now we have sixteen open source Deep Learning frameworks instead of just fifteen.

Spark Summit Europe Roundup

The 2015 Spark Summit Europe met in Amsterdam October 27-29.  Here is a roundup of the presentations, organized by subject areas.   I’ve omitted a few less interesting presentations, including some advertorials from sponsors.

State of Spark

— In his keynoter, Matei Zaharia recaps findings from Databricks’ Spark user survey, notes growth in summit attendance, meetup membership and contributor headcount.  (Video here). Enhancements expected for Spark 1.6:

  • Dataset API
  • DataFrame integration for GraphX, Streaming
  • Project Tungsten: faster in-memory caching, SSD storage, improved code generation
  • Additional data sources for Streaming

— Databricks co-founder Reynold Xin recaps the last twelve months of Spark development.  New user-facing developments in the past twelve months include:

  • DataFrames
  • Data source API
  • R binding and machine learning pipelines

Back-end developments include:

  • Project Tungsten
  • Sort-based shuffle
  • Netty-based network

Of these, Xin covers DataFrames and Project Tungsten in some detail.  Looking ahead, Xin discusses the Dataset API, Streaming DataFrames and additional Project Tungsten work.  Video here.

Getting Into Production

— Databricks engineer and Spark committer Aaron Davidson summarizes common issues in production and offers tips to avoid them.  Key issues: moving beyond Python performance; using Spark with R; network and CPU-bound workloads.  Video here.

— Tuplejump’s Evan Chan summarizes Spark deployment options and explains how to productionize Spark, with special attention to the Spark Job Server.  Video here.

— Spark committer and Databricks engineer Andrew Or explains how to use the Spark UI to visualize and debug performance issues.  Video here.

— Kostas Sakellis and Marcelo Vanzin of Cloudera provide a comprehensive overview of Spark security, covering encryption, authentication, delegation and authorization.  They tout Sentry, Cloudera’s preferred security platform.  Video here.

Spark for the Enterprise

— Revisting Matthew Glickman’s presentation at Spark Summit East earlier this year, Vinny Saulys reviews Spark’s impact at Goldman Sachs, noting the attractiveness of Spark’s APIs, in-memory processing and broad functionality.  He recaps Spark’s viral adoption within GS, and its broad use within the company’s data science toolkit.  His wish list for Spark: continued development of the DataFrame API; more built-in formulae; and a better IDE for Spark.  Video here.

— Alan Saldich summarizes Cloudera’s two years of experience working with Spark: a host of engineering contributions and 200+ customers (including Equifax, Barclays and a slide full of others).  Video here.  Key insights:

  • Prediction is the most popular use case
  • Hive is most frequently co-installed, followed by HBase, Impala and Solr.
  • Customers want security and performance comparable to leading relational databases combined with simplicity.

Data Sources and File Systems

— Stephan Kessler of SAP and Santiago Mola of Stratio explain Spark integration with SAP HANA Vora through the Data Sources API.  (Video unavailable).

— Tachyon Nexus’ Gene Pang offers an excellent overview of Tachyon’s memory-centric storage architecture and how to use Spark with Tachyon.  Video here.

Spark SQL and DataFrames

— Michael Armbrust, lead developer for Spark SQL, explains DataFrames.  Good intro for those unfamiliar with the feature.  Video here.

— For those who think you can’t do fast SQL without a Teradata box, Gianmario Spacagna showcases the Insight Engine, an application built on Spark.  More detail about the use case and solution here.  The application, which requires many very complex queries, runs 500 times faster on Spark than on Hive, and likely would not run at all on Teradata.  Video here.

— Informatica’s Kiran Lonikar summarizes a proposal to use GPUs to support columnar data frames.  Video here.

— Ema Orhian of Atigeo describes jaws, a restful data warehousing framework built on Spark SQL with Mesos and Tachyon support.  Video here.

Spark Streaming

— Helena Edelson, VP of Product Engineering at Tuplejump, offers a comprehensive overview of streaming analytics with Spark, Kafka, Cassandra and Akka.  Video here.

— Francois Garillot of Typesafe and Gerard Maas of virdata explain and demo Spark Streaming.    Video here.

— Iulian Dragos and Luc Bourlier explain how to leverage Mesos for Spark Streaming applications.  Video here.

Data Science and Machine Learning

— Apache Zeppelin creator and NFLabs co-founder Moon Soo Lee reviews the Data Science lifecycle, then demonstrates how Zeppelin supports development and collaboration through all phases of a project.  Video here.

— Alexander Ulanov, Senior Research Scientist at Hewlett-Packard Labs, describes his work with Deep Learning, building on MLLib’s multilayer perceptron capability.  Video here.

— Databricks’ Hossein Falaki offers an introduction to R’s strengths and weaknesses, then dives into SparkR.  He provides an overview of SparkR architecture and functionality, plus some pointers on mixing languages.  The SparkR roadmap, he notes, includes expanded MLLib functionality; UDF support; and a complete DataFrame API.  Finally, he demos SparkR and explains how to get started.  Video here.

— MLlib committer Joseph Bradley explains how to combine the strengths R, scikit-learn and MLlib.  Noting the strengths of R and scikit-learn libraries, he addresses the key question: how do you leverage software built to support single-machine workloads in a distributed computing environment?   Bradley demonstrates how to do this with Spark, using sentiment analysis as an example.  Video here.

— Natalino Busa of ING offers an introduction to real-time anomaly detection with Spark MLLib, Akka and Cassandra.  He describes different methods for anomaly detection, including distance-based and density-based techniques. Video here.

— Bitly’s Sarah Guido explains topic modeling, using Spark MLLib’s Latent Dirchlet Allocation.  Video here.

— Casey Stella describes using word2vec in MLLib to extract features from medical records for a Kaggle competition.  Video here.

— Piotr Dendek and Mateusz Fedoryszak of the University of Warsaw explain Random Ferns, a bagged form of Naive Bayes, for which they have developed a Spark package. Video here.

GeoSpatial Analytics

— Ram Sriharsha touts Magellan, an open source geospatial library that uses Spark as an engine.  Magellan, a Spark package, supports ESRI format files and GeoJSON; the developers aim to support the full suite of OpenGIS Simple Features for SQL.  Video here.

Use Cases and Applications

— Ion Stoica summarizes Databricks’ experience working with hundreds of companies, distills to two generic Spark use cases:  (1) the “Just-in-Time Data Warehouse”, bypassing IT bottlenecks inherent in conventional DW; (2) the unified compute engine, combining multiple frameworks in a single platform.  Video here.

— Apache committer and SKT engineer Yousun Jeong delivers a presentation documenting SKT’s Big Data architecture and a use case real-time analytics.  SKT needs to perform real-time analysis of the radio access network to improve utilization, as well as timely network quality assurance and fault analysis; the solution is a multi-layered appliance that combines Spark and other components with FPGA and Flash-based hardware acceleration.  Video here.

— Yahoo’s Ayman Farahat describes a collaborative filtering application built on Spark that generates 26 trillion recommendations.  Training time: 52 minutes; prediction time: 8 minutes.  Video here.

— Sujit Pal explains how Elsevier uses Spark together with Solr, OpenNLP to annotate documents at scale.  Elsevier has donated the application, called SoDA, back to open source.  Video here.

— Parkinson’s Disease affects one out of every 100 people over 60, and there is no cure.  Ido Karavany of Intel describes a project to use wearables to track the progression of the illness, using a complex stack including pebble, Android, IOS, play, Phoenix, HBase, Akka, Kafka, HDFS, MySQL and Spark, all running in AWS.   With Spark, the team runs complex computations daily on large data sets, and implements a rules engine to identify changes in patient behavior.  Video here.

— Paula Ta-Shma of IBM introduces a real-time routing use case from the Madrid bus system, then describes a solution that includes kafka, Secor, Swift, Parquet and elasticsearch for data collection; Spark SQL and MLLib for pattern learning; and a complex event processing engine for application in real time.  Video here.

Spark 1.4 Released

On June 11, the Spark team announced availability of Release 1.4.  More than 210 contributors from 70 different organizations contributed more than 1,000 patches.  Spark continues to expand its contributor base, the best measure of health for an open source project.

Screen Shot 2015-06-12 at 2.00.20 PM

Spark Core

The Spark team continues to improve Spark operability, performance and compatibility.  Key enhancements include:

  • The first phase in Project Tungsten performance improvements, a cache-friendly sort algorithm
  • Also for improved performance, serialized shuffle output
  • For the Spark UI, visualization for Spark DAGs and operational monitoring
  • A REST API for application information, such as job, stage, task and storage status
  • For Python users, support for Python 3.x, plus external spilling for Python groupByKey operations
  • Two YARN enhancements: support for YARN on EC2 and security for long-running YARN applications
  • Two Mesos enhancements: Docker support and cluster mode.

DataFrames and SQL

This release includes extensions of analytic functions for DataFrames, operational utilities for Spark SQL and support for ORCFile format.

A complete list of enhancements to the DataFrame API is here.

R Interface

AMPLab released a developer version of SparkR in January 2014.  In June 2014, Alteryx and Databricks announced a partnership to lead development of this component.  In March, 2015, SparkR officially merged into Spark.

SparkR offers an interface to use Apache Spark from R.  In Spark 1.4, SparkR supports operations like selection, filtering and aggregation on large datasets.  It’s important to note that as of this release SparkR does not support an interface to MLLib, Streaming or GraphX.

Machine Learning

In Spark 1.4, ML pipelines graduate from alpha release, add feature transformations (Vector Assembler, String Indexer, Bucketizer etc.) and a Python API.  Additional enhancements to ML include:

There appears to be an effort under way to rebuild MLLib’s supervised learning algorithms in ML.

Enhancements to MLLib include:

There is a single enhancement to GraphX in Spark 1.4, a personalized PageRank.  Spark’s graph analytics capabilities are comparatively static.

Streaming

The enhancements to Spark Streaming include improvements to the UI plus enhanced support for Kafka and Kinesis and a pluggable interface for write ahead logs.  Enhanced support for Kafka includes better error reporting, support for Kafka 0.8.2.1 and Kafka with Scala 2.11, input rate tracking and a Python API for Kakfa direct mode.

Spark Summit East: A Report (Updated)

Updated with links to slides where available.  Some links are broken, conference organizers have been notified.

Spark Summit East 2015 met on March 18 and 19 at the Sheraton Times Square in New York City.  Conference organizers announced another sellout (like the last two Spark Summits on the West Coast).

Competition for speaking slots at Spark events is heating up.  There were 170 submissions for 30 speaking slots at this event, compared to 85 submissions for 50 slots at Spark Summit 2014.  Compared to the last Spark Summit, presentations in the Applications Track, which I attended, were more polished, and demonstrate real progress in putting Spark to work.

The “father” of Spark, Matei Zaharia, kicked off the conference with a review of Spark progress in 2014 and planned enhancements for 2015.  Highlights of 2014 include:

  • Growth in contributors, from 150 to 500
  • Growth in the code base, from 190K lines to 370K lines
  • More than 500 known production instances at the close of 2014

Spark remains the most active project in the Hadoop ecosystem.

Also, in 2014, a team at Databricks smashed the Daytona GreySort record for petabyte-scale sorting.  The previous record, set in 2013, used MapReduce running on 2,100 machines to complete the task in 72 minutes.  The new record, set by Databricks with Spark running in the cloud, used 207 machines to complete the task in 23 minutes.

Key enhancements projected for 2015 include:

  • DataFrames, which are similar to frames in R, already released in Spark 1.3
  • R interface, which currently exists as SparkR, an independent project, targeted to be merged into Spark 1.4 in June
  • Enhancements to machine learning pipelines, which are sequences of tasks linked together into a process
  • Continued expansion of smart interfaces to external data sources, pushing logic into the sources
  • Spark packages — a repository for third-party packages (comparable to CRAN)

Databricks CEO Ion Stoica followed with a pitch for Databricks Cloud, which included brief testimonials from myfitnesspal, Automatic, Zoomdata, Uncharted Software and Tresata.

Additional keynoters included Brian Schimpf of Palantir, Matthew Glickman of Goldman Sachs and Peter Wang of Continuum Analytics.

Spark contributors presented detailed views on the current state of Spark:

  • Michael Armbrust, Spark SQL lead developer presented on the new DataFrames API and other enhancements to Spark SQL.
  • Tathagata Das delivered a talk on the current state and future of Spark Streaming.
  • Joseph Bradley covered MLLib, focusing on the Pipelines capability added in Spark 1.2
  • Ankur Dave offered an overview of GraphX, Spark’s graph engine.

Several observations from the Applications track:

(1) Geospatial applications had a strong presence.

  • Automatic, Tresata and Uncharted all showed live demonstrations of marketable products with geospatial components running on Spark
  • Mansour Raad of ESRI followed his boffo performance at Strata/Hadoop World last October with a virtuoso demonstration of Spark with massive spatial and temporal datasets and the ESRI open source GIS stack

(2) Spark provides a great platform for recommendation engines.

  • Comcast uses Spark to serve personalized recommendations based on analysis of billions of machine-generated events
  • Gilt Groupe uses Spark for a similar real-time application supporting flash sale events, where products are available for a limited time and in limited quantities
  • Leah McGuire of Salesforce described her work building a recommendation system using Spark

(3) Spark is gaining credibility in retail banking.

  • Sandy Ryza of Cloudera presented on Value At Risk (VAR) computations in Spark, a critical element in Basel reporting and stress testing
  • Startup Tresata demonstrated its application for Anti Money Laundering, which is built on a social graph built in Spark

(4) Spark has traction in the life sciences

  • Jeremy Freeman of HHMI Janelia Research Center, a regular presenter at Spark Summits, covered Spark’s unique capability for streaming machine learning.
  • David Tester of Novartis presented plans to build a trillion-edge graph for genomic integration
  • Timothy Danforth of Berkeley’s AMPLab delivered a presentation on next-generation genomics with Spark and ADAM
  • Kevin Mader of ETH Zurich spoke about turning big hairy 3D images into simple, robust, reproducible numbers without resorting to black boxes or magic

Also in the applications track: presenters from Baidu, myfitnesspal and Shopify.

Apache Spark is 1.0

Today, the Spark project announced availability of Apache Spark 1.0.0, the first major release since the Apache Foundation named Spark a top-level project.  (Additional announcements here, here and here). With 117 contributors, Spark continues to build critical mass and engagement in the data science community.

Features of the new release include:

  • API stability
  • Integration with YARN security
  • Operational and packaging improvements
  • Spark SQL (Alpha)
  • MLLib enhancements, including
    • Support for sparse feature vectors
    • Scalable decision trees for classification and regression
    • Distributed SVD and PCA
    • Model evaluation functions
    • L-BFGS optimization primitive
  • GraphX enhancements, including performance improvements in graph loading, edge reversal and neighborhood computation
  • Streaming enhancements, including optimized performance for stateful stream transformations, improved Flume support and automated state cleanup for long-running jobs
  • Extended Java and Python support
  • Significant improvements to documentation

…and many small improvements, documented in the Release Notes.

For more information on Spark, read this backgrounder.