Your mission-critical projects need complex event processing, realtime management and monitoring. Talend 5.4 (released in December 2013, https://www.talend.com) offers a great new feature: Talend Event Logging. It allows logging, processing and monitoring of all technical events and business data. In this article, I will focus on how to process, filter and monitor business data. You can find more details about monitoring technical events and logs (e.g. OSGi events of the ESB container) in Talend’s documentation (www.help.talend.com). This new feature is very powerful, but also extendable to fit custom requirements. You can solve and monitor much more complex scenarios than the one I describe here. Talend Event Logging with Logstash, ElasticSearch and Kibana First, let’s take a look at the components / projects which are integrated and extended into Talend’s products for implementing this new feature. logstash (http://logstash.net) is a tool for managing events and logs. You can use it to collect different (distributed) logs, parse them, and store them for later use. Speaking of searching, logstash comes with a simple, but fine web interface for searching and drilling into all of your logs. It uses ElasticSearch under the hood. So, you can easily query through all your logs for specific errors or business analytics (e.g. searching for all lines matching an unique order id). Additional to the pure collection of events, the Event Logging feature supports custom processing (e.g. custom filtering, customer data enrichment/reduction), aggregation, signing and also server side custom pre- and post-processing of events - e.g. to send them to an intrusion detection system or to any other kind of potential higher level log processing /management system. Kibana (http://www.elasticsearch.org/overview/kibana) is a browser based analytics and search interface to logstash and other timestamped data sets stored in ElasticSearch. Kibana strives to be easy to get started with, while also being flexible and powerful, just like logstash and ElasticSearch. Main difference to logstash is a much more powerful HTML5 based web interface. You can • use multiple concurrent search inputs • highlight to drill down bar charts • create line charts, stacked, unstacked, filled or unfilled, with or without points • create Pie and donut charts that compare top terms or the results of multiple queries • create custom dashboards with multiple charts • and much more… Therefore, logstash, Elasticsearch and Kibana are a perfect combination. You can use Kibana to analyze and monitor your data as you do with logstash, however, Kibana’s web interface is much more powerful and comfortable than logstash. There is a great book about logstash: “The logstash book” – for just 9.99 USD. I can really recommend this book for getting started: http://www.logstashbook.com/. For Elasticsearch, you can find several books on Amazon. Unfortunately, Kibana has no good and extensive documentation yet. I heard from its developers that this topic is addressed for Q1 2014. Integration of Event Logging and Monitoring into Talend’s Unified Platform Talend 5.4 has integrated logstash and Kibana into its Unified Platform. Talend Administration Center (TAC) is Talend’s central web application for management and monitoring. It got a new logging view: Here, you can use very flexible and powerful realtime search capabilities of Elasticsearch. Some dashboards are available by default. You can also create your own custom dashboards easily within this site thanks to Kibana. Many panel types are available, such as pie, histogram, table, hits or trends. You can analyze every technical event or business data down to the message level: By default, you see fields such as message, source, timestamp, type, and others. Of course, you can also add custom fields suitable for your business case. This way, you have a central monitoring capability which allows analyzing data on distributed clusters easily. Under the hood, this data comes from logstash. Many alternative inputs are available for logstash, such as log4j input or tcp input. For business data, I often use file input (http://logstash.net/docs/1.2.2/inputs/file) to analyze files such as CSV. Adding new inputs is very simple. You just have to add an input to your logstash configuration file of your logstash server. As I mentioned already, all this is integrated into the Talend’s Unified Platform: In this example, there are three log4j inputs and one file input. I use the file input to analyze text files in a specific directory. In this case, output is an embedded Elasticsearch instance. In production, you should use an external Elasticsearch cluster, of course. Processing such as filtering can also be configured in this file. Thus, you can process, analyze and monitor all your different inputs within one central monitoring application thanks to Kibana. Building Talend Integration Jobs, Routes and Web Services As mentioned before, you can analyze almost all data with logstash, Elasticsearch and Kibana. It does not matter if your input is technical events from a container (e.g. OSGi events) or any business data such as CSV files, log4j logs, or something else. Talend implicitly supports technical events which are created by the ESB container, by MDM, etc. However, you can also add your custom business data from your Talend jobs (integration perspective), SOAP / REST Web Services (integration perspective) or Talend routes (mediation perspective), easily: This is just one example (part of Talend’s DI demos which are included in every DI installation). The job generates some random data and stores it to a CSV file. You just have to add the configured file or directory of tFileOutputDelimited to your logstash configuration using file input (with wild cards for more complex scenarios). That’s it. You can now monitor and analyze the business data in realtime. This example showed a Talend DI Job (i.e. ETL job). However, you can also monitor your business data from SOAP / REST Web Services or Talend Routes the same way. Conclusion Your mission-critical projects need management and monitoring. Today, this is not just possible with complex and expensive tools of large vendors, but also with Talend’s Unified Platform products such as Talend ESB or Talend MDM. Under the hood, Talend integrates and extends widely used open source products: logstash, Elasticsearch and Kibana. Have fun with Talend 5.4’s new event logging and monitoring features… Best regards, Kai Wähner (@KaiWaehner) CONTENT FROM MY BLOG: http://www.kai-waehner.de/blog/2013/12/17/realtime-event-logging-complex-event-processing-cep-and-monitoring-with-talends-unified-platform-5-4-di-esb-dq-mdm-bpm-using-elasticsearch-logstash-and-kibana/

Editor's note: Be sure to check out part 2 as well. This is the fourth of an introductory series of blogs on Apache HBase. In the third part, we saw a high level view of HBase architecture . In this part, we'll use the HBase Java API to create tables, insert new data, and retrieve data by row key. We'll also see how to setup a basic table scan which restricts the columns retrieved and also uses a filter to page the results. Having just learned about HBase high-level architecture, now let's look at the Java client API since it is the way your applications interact with HBase. As mentioned earlier you can also interact with HBase via several flavors of RPC technologies like Apache Thrift plus a REST gateway, but we're going to concentrate on the native Java API. The client APIs provide both DDL (data definition language) and DML (data manipulation language) semantics very much like what you find in SQL for relational databases. Suppose we are going to store information about people in HBase, and we want to start by creating a new table. The following listing shows how to create a new table using the HBaseAdmin class. Configuration conf = HBaseConfiguration.create(); HBaseAdmin admin = new HBaseAdmin(conf); HTableDescriptor tableDescriptor = new HTableDescriptor(TableName.valueOf("people")); tableDescriptor.addFamily(new HColumnDescriptor("personal")); tableDescriptor.addFamily(new HColumnDescriptor("contactinfo")); tableDescriptor.addFamily(new HColumnDescriptor("creditcard")); admin.createTable(tableDescriptor); The people table defined in preceding listing contains three column families: personal, contactinfo, and creditcard. To create a table you create an HTableDescriptor and add one or more column families by adding HColumnDescriptor objects. You then call createTable to create the table. Now we have a table, so let's add some data. The next listing shows how to use the Put class to insert data on John Doe, specifically his name and email address (omitting proper error handling for brevity). Configuration conf = HBaseConfiguration.create(); HTable table = new HTable(conf, "people"); Put put = new Put(Bytes.toBytes("doe-john-m-12345")); put.add(Bytes.toBytes("personal"), Bytes.toBytes("givenName"), Bytes.toBytes("John")); put.add(Bytes.toBytes("personal"), Bytes.toBytes("mi"), Bytes.toBytes("M")); put.add(Bytes.toBytes("personal"), Bytes.toBytes("surame"), Bytes.toBytes("Doe")); put.add(Bytes.toBytes("contactinfo"), Bytes.toBytes("email"), Bytes.toBytes("[email protected]")); table.put(put); table.flushCommits(); table.close(); In the above listing we instantiate a Put providing the unique row key to the constructor. We then add values, which must include the column family, column qualifier, and the value all as byte arrays. As you probably noticed, the HBase API's utility Bytes class is used a lot; it provides methods to convert to and from byte[] for primitive types and strings. (Adding a static import for the toBytes() method would cut out a lot of boilerplate code.) We then put the data into the table, flush the commits to ensure locally buffered changes take effect, and finally close the table. Updating data is also done via the Put class in exactly the same manner as just shown in the prior listing. Unlike relational databases in which updates must update entire rows even if only one column changed, if you only need to update a single column then that's all you specify in the Put and HBase will only update that column. There is also a checkAndPut operation which is essentially a form of optimistic concurrency control - the operation will only put the new data if the current values are what the client says they should be. Retrieving the row we just created is accomplished using the Get class, as shown in the next listing. (From this point forward, listings will omit the boilerplate code to create a configuration, instantiate the HTable, and the flush and close calls.) Get get = new Get(Bytes.toBytes("doe-john-m-12345")); get.addFamily(Bytes.toBytes("personal")); get.setMaxVersions(3); Result result = table.get(get); The code in the previous listing instantiates a Get instance supplying the row key we want to find. Next we use addFamily to instruct HBase that we only need data from the personal column family, which also cuts down the amount of work HBase must do when reading information from disk. We also specify that we'd like up to three versions of each column in our result, perhaps so we can list historical values of each column. Finally, calling get returns a Result instance which can then be used to inspect all the column values returned. In many cases you need to find more than one row. HBase lets you do this by scanning rows, as shown in the second part which showed using a scan in the HBase shell session. The corresponding class is the Scan class. You can specify various options, such as the start and ending row key to scan, which columns and column families to include and the maximum versions to retrieve. You can also add filters, which allow you to implement custom filtering logic to further restrict which rows and columns are returned. A common use case for filters is pagination. For example, we might want to scan through all people whose last name is Smith one page (e.g. 25 people) at a time. The next listing shows how to perform a basic scan. Scan scan = new Scan(Bytes.toBytes("smith-")); scan.addColumn(Bytes.toBytes("personal"), Bytes.toBytes("givenName")); scan.addColumn(Bytes.toBytes("contactinfo"), Bytes.toBytes("email")); scan.setFilter(new PageFilter(25)); ResultScanner scanner = table.getScanner(scan); for (Result result : scanner) { // ... } In the above listing we create a new Scan that starts from the row key smith- and we then use addColumn to restrict the columns returned (thus reducing the amount of disk transfer HBase must perform) to personal:givenName and contactinfo:email. A PageFilter is set on the scan to limit the number of rows scanned to 25. (An alternative to using the page filter would be to specify a stop row key when constructing the Scan.) We then get a ResultScanner for the Scan just created, and loop through the results performing whatever actions are necessary. Since the only method in HBase to retrieve multiple rows of data is scanning by sorted row keys, how you design the row key values is very important. We'll come back to this topic later. You can also delete data in HBase using the Delete class, analogous to the Put class to delete all columns in a row (thus deleting the row itself), delete column families, delete columns, or some combination of those. Connection Handling In the above examples not much attention was paid to connection handling and RPCs (remote procedure calls). HBase provides the HConnection class which provides functionality similar to connection pool classes to share connections, for example you use the getTable() method to get a reference to an HTable instance. There is also an HConnectionManager class which is how you get instances of HConnection. Similar to avoiding network round trips in web applications, effectively managing the number of RPCs and amount of data returned when using HBase is important, and something to consider when writing HBase applications. Conclusion to Part 4 In this part we used the HBase Java API to create a people table, insert a new person, and find the newly inserted person information. We also used the Scan class to scan the people table for people with last name "Smith" and showed how to restrict the data retrieved and finally how to use a filter to limit the number of results. In the next part, we'll learn how to deal with the absence of SQL and relations when modeling schemas in HBase. References HBase web site, http://hbase.apache.org/ HBase wiki, http://wiki.apache.org/hadoop/Hbase HBase Reference Guide http://hbase.apache.org/book/book.html HBase: The Definitive Guide, http://bit.ly/hbase-definitive-guide Google Bigtable Paper, http://labs.google.com/papers/bigtable.html Hadoop web site, http://hadoop.apache.org/ Hadoop: The Definitive Guide, http://bit.ly/hadoop-definitive-guide Fallacies of Distributed Computing, http://en.wikipedia.org/wiki/Fallacies_of_Distributed_Computing HBase lightning talk slides, http://www.slideshare.net/scottleber/hbase-lightningtalk Sample code, https://github.com/sleberknight/basic-hbase-examples

One thing that Big Data certainly made happen is that it brought the database/infrastructure community and the data analysis/statistics/machine learning communities closer together. As always, each community had it’s own set of models, methods, and ideas about how to structure and interpret the world. You can still tell these differences when looking at current Big Data projects, and I think it’s important to be aware of the distinctions in order to better understand the relationships between different projects. Because, let’s face it, every project claims to re-invent Big Data. Hadoop and MapReduce being something like the founding fathers of Big Data, other’s projects have since appeared. Most notably, there are stream processing projects like Twitter’s Storm who move from batch-oriented processing to event-based processing which is more suited for real-time, low-latency processing. Spark is yet something different, a bit like Hadoop, but puts greater emphasis on iterative algorithms, and in-memory processing to achieve that landmark “100x faster than Hadoop” every current project seems to need to sport. Twitter’s summingbird project tries to bridge the gap between MapReduce and stream processing by providing us with a high-level set of operators which can then either run on MapReduce or Storm. However, both Spark or summingbird leave me sort of flat because you can see that they come from a database background, which means that there will still be a considerable gap to serious machine learning. So, what exactly is the difference? In the end, it’s the difference between relational and linear algebra. In the database world, you model relationships between objects, which you encode in tables, and foreign keys to link up entries between different tables. Probably the most important insight of the database world was to develop a query language, a declarative description of what you want to extract from your database, leaving the optimization of the query and the exact details of how to perform them efficiently to the database guys. The machine learning community, on the other hand, has its roots in linear algebra and probability theory. Objects are usually encoded as a feature vector, that is, a list of numbers describing different properties of an object. Data is often collected in matrices where each row corresponds to an object, and each column to a feature, not much unlike a table in a database. However, the operations you perform in order to do data analysis are quite different from the data base world. Take something as basic as linear regression: your try to learn a linear function f(x)=di=1wixi in a d-dimensional space (that is, where your objects are described by a d-dimensional vector) given n examples Xi, and Yi, where Xi are the features describing your objects and Yi is the real number you attach to Xi. One way to “learn” w is to tune it such that the quadratic error on the training examples is minimal. The solution can be written in closed form as w=(XXT)−1XY where X is the matrix built from the Xi (putting the Xi in the columns of X), and Y is the vector of outputs Yi. In order to solve this, you need to solve the linear equation (XXT)w=XY which can be done by one of a large number of algorithms, starting with Gaussian elimination, which you’ve probably learned in your undergrad studies, or the conjugate gradient algorithm, or by first computing a Cholesky decomposition. All of these algorithms have in common that they are iterative. They go through a number of operations, for example O(d3) for the Gaussian elimination case. They also need to store intermediate results. Gaussian elimination and Cholesky decomposition have rather elementary operations acting on individual entries, while the conjugate gradient algorithm performs a matrix-vector multiplication in each iteration. Most importantly, these algorithms can only be expressed very badly in SQL! It’s certainly not impossible, but you’d need to store your data in much different ways than you would in idiomatic database usage. So, it’s not about whether or not your framework can support iterative algorithms without significant latency, it’s about understanding that joins, group bys, and count() won’t get you far, but you need scalar products, matrix-vector and matrix-matrix multiplications. You don’t need indices for most ML algorithms, maybe except for being able to quickly find the k-nearest neighbors, because most algorithms tend to either take in the whole data set in each iteration or otherwise stream the whole set by some model which is iteratively updated like in stochastic gradient descent. I’m not sure projects like Spark or Stratosphere have fully grasped the significance of this yet. Database infrastructure-inspired Big Data has it’s place when it comes to extracting and preprocessing data, but eventually, you move from database land to machine learning land, which invariably means linear algebra land (or probability theory land, which often also reduces to linear algebra like computations). What often happens today is that you either painstakingly have to break down your linear algebra into MapReduce jobs, or you actively look for algorithms which fit the database view better. I think we’re still at the beginning of what is possible. Or, to be a bit more aggressive, claims that existing (infrastructure, database, parallelism inspired) frameworks provide you with sophistic data analytics are widely exaggerated. They take care of a very important problem by giving you a reliable infrastructure to scale your data analysis code, but there’s still a lot of work that needs to be done on your side. High-level DSLs like Apache Hive or Pig are a first step in this direction but still too much rooted in the database world IMHO. In summary, one should be aware of the difference between a framework which mostly is concerned with scaling and a tool which actually provides some piece of data analysis. And even if it comes with basic database-like analytics mechanisms, there is still a long way to go to do some serious data science. That’s why we’re also thinking that streamdrill occupies an interesting spot, because it is a bit of infrastructure, allowing you to process a serious amount of event data, but it also provides valuable analysis based on algorithms you wouldn’t want to implement yourself, even if you had some Big Data framework like Hadoop at hand. That’s an interesting direction I also would like to see more of in the future. Note: Just saw that Spark has a logistic regression example on their landing page. Well, doing matrix operations explicitly via map() on collections doesn’t count in my view ;)

ElasticSearch provides Java API, thus it executes all operations asynchronously by using client object.

Learn about ElasticSearch, an open source tool developed with Java. It is a Lucene-based, scalable, full-text search engine, and a data analysis tool.

Originally posted on the SpringSource blog by Janne Valkealahti We're super excited to let the cat out of the bag and release support for writing YARN based applications as part of the Spring for Apache Hadoop 2.0 M1 release. In this blog post I’ll introduce you to YARN, what you can do with it, and how Spring simplifies the development of YARN based applications. If you have been following the Hadoop community over the past year or two, you’ve probably seen a lot of discussions around YARN and the next version of Hadoop's MapReduce called MapReduce v2. YARN (Yet Another Resource Negotiator) is a component of the MapReduce project created to overcome some performance issues in Hadoop's original design. The fundamental idea of MapReduce v2 is to split the functionalities of the JobTracker, Resource Management and Job Scheduling/Monitoring, into separate daemons. The idea is to have a global Resource Manager (RM) and a per-application Application Master (AM). A generic diagram for YARN component dependencies can be found from YARN architecture. MapReduce Version 2 is an application running on top of YARN. It is also possible to make similar custom YARN based application which have nothing to do with MapReduce, it is simply running YARN application. However, writing a custom YARN based application is difficult. The YARN APIs are low-level infrastructure APIs and not developer APIs. Take a look at thedocumentation for writing a YARN application to get an idea of what is involved. Starting with the 2.0 version, Spring for Apache Hadoop introduces the Spring YARN sub-project to provide support for building Spring based YARN applications. This support for YARN steps in by trying to make development easier. “Spring handles the infrastructure so you can focus on your application” applies to writing Hadoop applications as well as other types of Java applications. Spring’s YARN support also makes it easier to test your YARN application. With Spring’s YARN support, you're going to use all familiar concepts of Spring Framework itself, including configuration and generally speaking what you can do in your application. At a high level, Spring YARN provides three different components, YarnClient, YarnAppmaster andYarnContainer which together can be called a Spring YARN Application. We provide default implementations for all components while still giving the end user an option to customize as much as he or she wants. Lets take a quick look at a very simplistic Spring YARN application which runs some custom code in a Hadoop cluster. The YarnClient is used to communicate with YARN's Resource Manager. This provides management actions like submitting new application instances, listing applications and killing running applications. When submitting applications from the YarnClient, the main concerns relate to how the Application Master is configured and launched. Both the YarnAppmaster andYarnContainer share the same common launch context config logic so you'll see a lot of similarities in YarnClient and YarnAppmaster configuration. Similar to how the YarnClient will define the launch context for the YarnAppmaster, the YarnAppmaster defines the launch context for the YarnContainer. The Launch context defines the commands to start the container, localized files, command line parameters, environment variables and resource limits(memory, cpu). The YarnContainer is a worker that does the heavy lifting of what a YARN application will actually do. The YarnAppmaster is communicating with YARN Resource Manager and starts and stops YarnContainers accordingly. You can create a Spring application that launches an ApplicationMaster by using the YARN XML namespace to define a Spring Application Context. Context configuration for YarnClient defines the launch context for YarnAppmaster. This includes resources and libraries needed by YarnAppmaster and its environment settings. An example of this is shown below. Note: Future releases will provide a Java based API for configuration, similar to what is done in Spring Security 3.2. The purpose of YarnAppmaster is to control the instance of a running application. TheYarnAppmaster is responsible for controlling the lifecycle of all its YarnContainers, the whole running application after the application is submitted, as well as itself. The example above is defining a context configuration for the YarnAppmaster. Similar to what we saw in YarnClient configuration, we define local resources for the YarnContainer and its environment. The classpath setting picks up hadoop jars as well as your own application jars in default locations, change the setting if you want to use non-default directories. Also within theYarnAppmaster we define components handling the container allocation and bootstrapping. Allocator component is interacting with YARN resource manager handling the resource scheduling. Runner component is responsible for bootstrapping of allocated containers. Above example defines a simple YarnContainer context configuration. To implement the functionality of the container, you implement the interface YarnContainer. The YarnContainer interface is similar to Java’s Runnable interface, its has a run() method, as well as two additional methods related to getting environment and command line information. Below is a simple hello world application that will be run inside of a YARN container: public class MyCustomYarnContainer implements YarnContainer { private static final Log log = LogFactory.getLog(MyCustomYarnContainer.class); @Override public void run() { log.info("Hello from MyCustomYarnContainer"); } @Override public void setEnvironment(Map environment) {} @Override public void setParameters(Properties parameters) {} } We just showed the configuration of a Spring YARN Application and the core application logic so what remains is how to bootstrap the application to run inside the Hadoop cluster. The utility class, CommandLineClientRunner provides this functionality. You can you use CommandLineClientRunner either manually from a command line or use it from your own code. # java -cp org.springframework.yarn.client.CommandLineClientRunner application-context.xml yarnClient -submit A Spring YARN Application is packaged into a jar file which then can be transferred into HDFS with the rest of the dependencies. A YarnClient can transfer all needed libraries into HDFS during the application submit process but generally speaking it is more advisable to do this manually in order to avoid unnecessary network I/O. Your application wont change until new version is created so it can be copied into HDFS prior the first application submit. You can i.e. use Hadoop's hdfs dfs -copyFromLocal command. Below you can see an example of a typical project setup. src/main/java/org/example/MyCustomYarnContainer.java src/main/resources/application-context.xml src/main/resources/appmaster-context.xml src/main/resources/container-context.xml As a wild guess, we'll make a bet that you have now figured out that you are not actually configuring YARN, instead you are configuring Spring Application Contexts for all three components, YarnClient, YarnAppmaster and YarnContainer. We have just scratched the surface of what we can do with Spring YARN. While we’re preparing more blog posts, go ahead and check existing samples in GitHub. Basically, to reflect the concepts we described in this blog post, see the multi-context example in our samples repository. Future blog posts will cover topics like Unit Testing and more advanced YARN application development.

introduction openstack is one of the most popular open source cloud computing projects to provide infrastructure as a service solution. its key components are compute (nova), networking (neutron, formerly known as quantum), storage (object and block storage, swift and cinder, respectively), openstack dashboard (horizon), identity service (keystone) and image service (glance). there are other official incubated projects like metering (celiometer) and orchestration and service definition (heat). savanna is a hadoop as a service for openstack introduced by mirantis . it is still in an early phase (version .02 was released in summer 2013) and according to its roadmap version 1.0 is targeted for official openstack incubation. in principle, heat also could be used for hadoop cluster provisioning but savanna is especially tuned for providing hadoop-specific api functionality while heat is meant to be used for generic purposes. savanna architecture savanna is integrated with the core openstack components such as keystone, nova, glance, swift and horizon. it has a rest api that supports the hadoop cluster provisioning steps. savanna api is implemented as a wsgi server that, by default, listens to port 8386. in addition, savanna can also be integrated with horizon, the openstack dashboard to create a hadoop cluster from the management console. savanna also comes with a vanilla plugin that deploys a hadoop cluster image. the standard out-of-the-box vanilla plugin supports hadoop 1.1.2 version. installing savanna the simplest option to try out savanna is to use devstack in a virtual machine. i was using an ubuntu 12.04 virtual instance in my tests. in that environment we need to execute the following commands to install devstack and savanna api: $ sudo apt-get install git-core $ git clone https://github.com/openstack-dev/devstack.git $ vi localrc # edit localrc admin_password=nova mysql_password=nova rabbit_password=nova service_password=$admin_password service_token=nova # enable swift enabled_services+=,swift swift_hash=66a3d6b56c1f479c8b4e70ab5c2000f5 swift_replicas=1 swift_data_dir=$dest/data # force checkout prerequsites # force_prereq=1 # keystone is now configured by default to use pki as the token format which produces huge tokens. # set uuid as keystone token format which is much shorter and easier to work with. keystone_token_format=uuid # change the floating_range to whatever ips vm is working in. # in nat mode it is subnet vmware fusion provides, in bridged mode it is your local network. floating_range=192.168.55.224/27 # enable auto assignment of floating ips. by default savanna expects this setting to be enabled extra_opts=(auto_assign_floating_ip=true) # enable logging screen_logdir=$dest/logs/screen $ ./stack.sh # this will take a while to execute $ sudo apt-get install python-setuptools python-virtualenv python-dev $ virtualenv savanna-venv $ savanna-venv/bin/pip install savanna $ mkdir savanna-venv/etc $ cp savanna-venv/share/savanna/savanna.conf.sample savanna-venv/etc/savanna.conf # to start savanna api: $ savanna-venv/bin/python savanna-venv/bin/savanna-api --config-file savanna-venv/etc/savanna.conf to install savanna ui integrated with horizon, we need to run the following commands: $ sudo pip install savanna-dashboard $ cd /opt/stack/horizon/openstack-dashboard $ vi settings.py horizon_config = { 'dashboards': ('nova', 'syspanel', 'settings', 'savanna'), installed_apps = ( 'savannadashboard', .... $ cd /opt/stack/horizon/openstack-dashboard/local $ vi local_settings.py savanna_url = 'http://localhost:8386/v1.0' $ sudo service apache2 restart provisioning a hadoop cluster as a first step, we need to configure keystone-related environment variables to get the authentication token: ubuntu@ip-10-59-33-68:~$ vi .bashrc $ export os_auth_url=http://127.0.0.1:5000/v2.0/ $ export os_tenant_name=admin $ export os_username=admin $ export os_password=nova ubuntu@ip-10-59-33-68:~$ source .bashrc ubuntu@ip-10-59-33-68:~$ ubuntu@ip-10-59-33-68:~$ env | grep os os_password=nova os_auth_url=http://127.0.0.1:5000/v2.0/ os_username=admin os_tenant_name=admin ubuntu@ip-10-59-33-68:~$ keystone token-get +-----------+----------------------------------+ | property | value | +-----------+----------------------------------+ | expires | 2013-08-09t20:31:12z | | id | bdb582c836e3474f979c5aa8f844c000 | | tenant_id | 2f46e214984f4990b9c39d9c6222f572 | | user_id | 077311b0a8304c8e86dc0dc168a67091 | +-----------+----------------------------------+ $ export auth_token="bdb582c836e3474f979c5aa8f844c000" $ export tenant_id="2f46e214984f4990b9c39d9c6222f572" then we need to create the glance image that we want to use for our hadoop cluster. in our example we have used mirantis's vanilla image but we can also build our own image: $ wget http://savanna-files.mirantis.com/savanna-0.2-vanilla-1.1.2-ubuntu-12.10.qcow2 $ glance image-create --name=savanna-0.2-vanilla-hadoop-ubuntu.qcow2 --disk-format=qcow2 --container-format=bare < ./savanna-0.2-vanilla-1.1.2-ubuntu-12.10.qcow2 ubuntu@ip-10-59-33-68:~/devstack$ glance image-list +--------------------------------------+-----------------------------------------+-------------+------------------+-----------+--------+ | id | name | disk format | container format | size | status | +--------------------------------------+-----------------------------------------+-------------+------------------+-----------+--------+ | d0d64f5c-9c15-4e7b-ad4c-13859eafa7b8 | cirros-0.3.1-x86_64-uec | ami | ami | 25165824 | active | | fee679ee-e0c0-447e-8ebd-028050b54af9 | cirros-0.3.1-x86_64-uec-kernel | aki | aki | 4955792 | active | | 1e52089b-930a-4dfc-b707-89b568d92e7e | cirros-0.3.1-x86_64-uec-ramdisk | ari | ari | 3714968 | active | | d28051e2-9ddd-45f0-9edc-8923db46fdf9 | savanna-0.2-vanilla-hadoop-ubuntu.qcow2 | qcow2 | bare | 551699456 | active | +--------------------------------------+-----------------------------------------+-------------+------------------+-----------+--------+ $ export image_id=d28051e2-9ddd-45f0-9edc-8923db46fdf9 then we have installed httpie , an open source http client that can be used to send rest requests to savanna api: $ sudo pip install httpie from now on we will use httpie to send savanna commands. we need to register the image with savanna: $ export savanna_url="http://localhost:8386/v1.0/$tenant_id" $ http post $savanna_url/images/$image_id x-auth-token:$auth_token username=ubuntu http/1.1 202 accepted content-length: 411 content-type: application/json date: thu, 08 aug 2013 21:28:07 gmt { "image": { "os-ext-img-size:size": 551699456, "created": "2013-08-08t21:05:55z", "description": "none", "id": "d28051e2-9ddd-45f0-9edc-8923db46fdf9", "metadata": { "_savanna_description": "none", "_savanna_username": "ubuntu" }, "mindisk": 0, "minram": 0, "name": "savanna-0.2-vanilla-hadoop-ubuntu.qcow2", "progress": 100, "status": "active", "tags": [], "updated": "2013-08-08t21:28:07z", "username": "ubuntu" } } $ http $savanna_url/images/$image_id/tag x-auth-token:$auth_token tags:='["vanilla", "1.1.2", "ubuntu"]' http/1.1 202 accepted content-length: 532 content-type: application/json date: thu, 08 aug 2013 21:29:25 gmt { "image": { "os-ext-img-size:size": 551699456, "created": "2013-08-08t21:05:55z", "description": "none", "id": "d28051e2-9ddd-45f0-9edc-8923db46fdf9", "metadata": { "_savanna_description": "none", "_savanna_tag_1.1.2": "true", "_savanna_tag_ubuntu": "true", "_savanna_tag_vanilla": "true", "_savanna_username": "ubuntu" }, "mindisk": 0, "minram": 0, "name": "savanna-0.2-vanilla-hadoop-ubuntu.qcow2", "progress": 100, "status": "active", "tags": [ "vanilla", "ubuntu", "1.1.2" ], "updated": "2013-08-08t21:29:25z", "username": "ubuntu" } } then we need to create a nodegroup templates (json files) that will be sent to savanna. there is one template for the master nodes ( namenode , jobtracker ) and another template for the worker nodes such as datanode and tasktracker . the hadoop version is 1.1.2. $ vi ng_master_template_create.json { "name": "test-master-tmpl", "flavor_id": "2", "plugin_name": "vanilla", "hadoop_version": "1.1.2", "node_processes": ["jobtracker", "namenode"] } $ vi ng_worker_template_create.json { "name": "test-worker-tmpl", "flavor_id": "2", "plugin_name": "vanilla", "hadoop_version": "1.1.2", "node_processes": ["tasktracker", "datanode"] } $ http $savanna_url/node-group-templates x-auth-token:$auth_token < ng_master_template_create.json http/1.1 202 accepted content-length: 387 content-type: application/json date: thu, 08 aug 2013 21:58:00 gmt { "node_group_template": { "created": "2013-08-08t21:58:00", "flavor_id": "2", "hadoop_version": "1.1.2", "id": "b3a79c88-b6fb-43d2-9a56-310218c66f7c", "name": "test-master-tmpl", "node_configs": {}, "node_processes": [ "jobtracker", "namenode" ], "plugin_name": "vanilla", "updated": "2013-08-08t21:58:00", "volume_mount_prefix": "/volumes/disk", "volumes_per_node": 0, "volumes_size": 10 } } $ http $savanna_url/node-group-templates x-auth-token:$auth_token < ng_worker_template_create.json http/1.1 202 accepted content-length: 388 content-type: application/json date: thu, 08 aug 2013 21:59:41 gmt { "node_group_template": { "created": "2013-08-08t21:59:41", "flavor_id": "2", "hadoop_version": "1.1.2", "id": "773b2cfb-1e05-46f4-923f-13edc7d6aac6", "name": "test-worker-tmpl", "node_configs": {}, "node_processes": [ "tasktracker", "datanode" ], "plugin_name": "vanilla", "updated": "2013-08-08t21:59:41", "volume_mount_prefix": "/volumes/disk", "volumes_per_node": 0, "volumes_size": 10 } } the next step is to define the cluster template: $ vi cluster_template_create.json { "name": "demo-cluster-template", "plugin_name": "vanilla", "hadoop_version": "1.1.2", "node_groups": [ { "name": "master", "node_group_template_id": "b3a79c88-b6fb-43d2-9a56-310218c66f7c", "count": 1 }, { "name": "workers", "node_group_template_id": "773b2cfb-1e05-46f4-923f-13edc7d6aac6", "count": 2 } ] } $ http $savanna_url/cluster-templates x-auth-token:$auth_token < cluster_template_create.json http/1.1 202 accepted content-length: 815 content-type: application/json date: fri, 09 aug 2013 07:04:24 gmt { "cluster_template": { "anti_affinity": [], "cluster_configs": {}, "created": "2013-08-09t07:04:24", "hadoop_version": "1.1.2", "id": "{ "name": "cluster-1", "plugin_name": "vanilla", "hadoop_version": "1.1.2", "cluster_template_id" : "64c4117b-acee-4da7-937b-cb964f0471a9", "user_keypair_id": "stack", "default_image_id": "3f9fc974-b484-4756-82a4-bff9e116919b" }", "name": "demo-cluster-template", "node_groups": [ { "count": 1, "flavor_id": "2", "name": "master", "node_configs": {}, "node_group_template_id": "b3a79c88-b6fb-43d2-9a56-310218c66f7c", "node_processes": [ "jobtracker", "namenode" ], "volume_mount_prefix": "/volumes/disk", "volumes_per_node": 0, "volumes_size": 10 }, { "count": 2, "flavor_id": "2", "name": "workers", "node_configs": {}, "node_group_template_id": "773b2cfb-1e05-46f4-923f-13edc7d6aac6", "node_processes": [ "tasktracker", "datanode" ], "volume_mount_prefix": "/volumes/disk", "volumes_per_node": 0, "volumes_size": 10 } ], "plugin_name": "vanilla", "updated": "2013-08-09t07:04:24" } } now we are ready to create the hadoop cluster: $ vi cluster_create.json { "name": "cluster-1", "plugin_name": "vanilla", "hadoop_version": "1.1.2", "cluster_template_id" : "64c4117b-acee-4da7-937b-cb964f0471a9", "user_keypair_id": "savanna", "default_image_id": "d28051e2-9ddd-45f0-9edc-8923db46fdf9" } $ http $savanna_url/clusters x-auth-token:$auth_token < cluster_create.json http/1.1 202 accepted content-length: 1153 content-type: application/json date: fri, 09 aug 2013 07:28:14 gmt { "cluster": { "anti_affinity": [], "cluster_configs": {}, "cluster_template_id": "64c4117b-acee-4da7-937b-cb964f0471a9", "created": "2013-08-09t07:28:14", "default_image_id": "d28051e2-9ddd-45f0-9edc-8923db46fdf9", "hadoop_version": "1.1.2", "id": "d919f1db-522f-45ab-aadd-c078ba3bb4e3", "info": {}, "name": "cluster-1", "node_groups": [ { "count": 1, "created": "2013-08-09t07:28:14", "flavor_id": "2", "instances": [], "name": "master", "node_configs": {}, "node_group_template_id": "b3a79c88-b6fb-43d2-9a56-310218c66f7c", "node_processes": [ "jobtracker", "namenode" ], "updated": "2013-08-09t07:28:14", "volume_mount_prefix": "/volumes/disk", "volumes_per_node": 0, "volumes_size": 10 }, { "count": 2, "created": "2013-08-09t07:28:14", "flavor_id": "2", "instances": [], "name": "workers", "node_configs": {}, "node_group_template_id": "773b2cfb-1e05-46f4-923f-13edc7d6aac6", "node_processes": [ "tasktracker", "datanode" ], "updated": "2013-08-09t07:28:14", "volume_mount_prefix": "/volumes/disk", "volumes_per_node": 0, "volumes_size": 10 } ], "plugin_name": "vanilla", "status": "validating", "updated": "2013-08-09t07:28:14", "user_keypair_id": "savanna" } } after a while we can run the nova command to check if the instances are created and running: $ nova list +--------------------------------------+-----------------------+--------+------------+-------------+----------------------------------+ | id | name | status | task state | power state | networks | +--------------------------------------+-----------------------+--------+------------+-------------+----------------------------------+ | 1a9f43bf-cddb-4556-877b-cc993730da88 | cluster-1-master-001 | active | none | running | private=10.0.0.2, 192.168.55.227 | | bb55f881-1f96-4669-a94a-58cbf4d88f39 | cluster-1-workers-001 | active | none | running | private=10.0.0.3, 192.168.55.226 | | 012a24e2-fa33-49f3-b051-9ee2690864df | cluster-1-workers-002 | active | none | running | private=10.0.0.4, 192.168.55.225 | +--------------------------------------+-----------------------+--------+------------+-------------+----------------------------------+ now we can log in to the hadoop master instance and run the required hadoop commands: $ ssh -i savanna.pem [email protected] $ sudo chmod 777 /usr/share/hadoop $ sudo su hadoop $ cd /usr/share/hadoop $ hadoop jar hadoop-example-1.1.2.jar pi 10 100 savanna ui via horizon in order to create nodegroup templates, cluster templates and the cluster itself we used a command line tool -- httpie -- to send rest api calls. the same functionality is also available via horizon, the standard openstack dashboard. first we need to register the image with savanna: then we need to create the nodegroup templates: after that we have to create the cluster template: and finally we have to create the cluster:

Dan McCreary and Ann Kelly, authors of 'Making Sense of NoSQL,' discuss the business drivers and motivations that make NoSQL so popular to organizations today.

In this blog post, I will show you how to "ETL" all kinds of data to Amazon’s cloud data warehouse Redshift wit Talend’s big data components. Let’s begin with a short introduction to Amazon Redshift (copied from website): "Amazon Redshift is [part of Amazon Web Services (AWS) and] a fast and powerful, fully managed, petabyte-scale data warehouse service in the cloud. With a few clicks in the AWS Management Console, customers can launch a Redshift cluster, starting with a few hundred gigabytes and scaling to a petabyte or more, for under $1,000 per terabyte per year. Traditional data warehouses require significant time and resource to administer, especially for large datasets. In addition, the financial cost associated with building, maintaining, and growing self-managed, on-premise data warehouses is very high. Amazon Redshift not only significantly lowers the cost of a data warehouse, but also makes it easy to analyze large amounts of data very quickly.“ Sounds interesting! And indeed, we already see companies using Talend’s Redshift connectors. From Talend perspective it is not much more than just another database. If you have ever used a Talend connector, you can integrate to Redshift within some minutes. In the next sections, I will describe all necessary steps and give some hints regarding configuration issues and performance improvements. Be aware: You need Talend Open Studio for Data Integration (Apache License, open source) or any Talend Enterprise Edition / Platform which contains the Cloud components to see and use Amazon Redshift connectors. The open source edition offers all connectors and functionality to integrate with Amazon Redshift. However, Enterprise versions offer some more features (e.g. versioning), comfort (e.g. wizards) and commercial support. Setup Amazon Redshift Setup of Amazon Redshift is very easy. Just follow Amazon‘s getting started guide: http://docs.aws.amazon.com/redshift/latest/gsg/welcome.html. Like every other AWS guide, it is very easy to understand and use. Be aware, that you just have to do step 1, 2 and 3 of the getting started guide for using it with Talend. Some hints: - Step 1 („before you begin“): Just sign up. Client tools and drivers are not necessary because they are already installed within Talend Studio. - Step 2 („launch a cluster“): Yes, please start your cluster! - Step 3(„authorize access“): If you are not sure what to do here, select Connection Type = CIDR/IP. Find out your IP address (http://whatismyipaddress.com) and enter it with „/32“ at the end. Example: „192.168.1.1/32“ Now you can connect to Amazon Redshift from your Talend Studio on your local computer. Step 4 (connect) and step 5 (create table, data, queries) are not necessary, this will be done from Talend Studio. Of course, you should not forget to delete your cluster (step 7) when you are done. Otherwise, you will pay for every hour, even if you do not access your DWH. Connect to Amazon Redshift from Talend Studio Create a new connection to Amazon Redshift database as you do with every other relational database. The easiest way is to use „DB Connection Wizard“ in metadata. Just enter your connection information and check if it works. You get all information about configuration from Amazon Web Console. The connection string looks something like this: „jdbc:paraccel://talend-demo-cluster.cp8t6c5.eu-west-1.redshift.amazonaws.com:5439/dev“ Next, right click on the created connection and select „retrieve schema“. „public“ is the default schema which you (have to) use. Now, you are ready to use this connection within Talend Jobs to write to Amazon Redshift and read from it. Create Talend Jobs (Write, Read, Delete) Amazon Redshift components work like any other Talend (relational) database components. Look at www.help.talend.com for more information if you have not used them before (or just try them out, they are very self-explanatory). You just have to drag&drop your connection from metadata . Afterwards, you can easily write data (tRedShiftOutput), read data (tRedshiftInput), or do any other queries such as delete or copy (tRedShiftRow). In the following job, I start with deleting all content in the Amazon Redshift table. Then, I read data from a MySQL table and insert it into an Amazon Redshift table. The table is created automatically (as I have configured it this way). After this subjob is finished, I read the data again, and store it to a CSV file (which is also created automatically). Of course, this is no business use case, but it shows how to use different Amazon Redshift components. Query Data from Amazon Redshift You can connect to Amazon Redshift directly from Talend Studio to explore and query data of the DWH. Thus, no other database tool is required. Just right click on your Amazon Redshift connection in metadata and select „edit queries“. Here you can define, execute and save SQL queries. Improve Performance Write performance of Amazon Redshift is relatively low compared to „classical“ relational databases (in your data center) as you have to upload all data into the cloud. Different alternatives exist to improve performance: - Bulk inserts: „Extended insert“ (in advanced settings) improves performance a lot, but still not to hyperspeed… Also, as it is bulk, you can just do inserts! It is not compatible to „rejects“ or „updates“ - AWS S3 and COPY command: S3 is Amazon’s „simple storage service“, a key-value store – also called NoSQL today – for storing very large objects. You can use Amazon Redshift’s COPY command (http://docs.aws.amazon.com/redshift/latest/dg/r_COPY.html) to transfer data from S3 to Amazon Redshift with good performance. Though, you still have to copy data to S3 before, same „cloud problem“ here. The COPY command can be used with tRedshiftRow, so no problem at all from Talend perspective. To transfer data to S3, you can either use the Talend S3 components from Talendforge, Talend’s open source community (http://www.talendforge.org/exchange), or use camel-s3, an Apache Camel component which is included in Talend ESB. The latter is an option, if you use Talend Data Services which combines Talend DI and Talend ESB in its unified platform. Summary You need not be a cloud or DWH expert, or an expert developer to integrate with Amazon’s cloud data warehouse Redshift. It is very easy with Talend’s integration solutions. Just drag&drop, configure, do some graphical mappings / transformations (if necessary), that’s it. Code is generated. Job runs. You can integrate Amazon Redshift almost as simple as any other relational database. Just be aware of some cloud specific security and performance issues. With Talend, you can easily „ETL“ all data from different sources to Redshift and store it there for under $1,000 per terabyte per year – even with the open source version! Best regards, Kai Wähner (Contact and feedback via @KaiWaehner, www.kai-waehner.de, LinkedIn / Xing) This is content from my blog: http://www.kai-waehner.de/blog/2013/06/26/integration-of-amazon-redshift-cloud-data-warehouse-aws-saas-dwh-with-talend-data-integration-di-big-data-bd-enterprise-service-bus-esb/

Elasticsearch is a flexible and powerful open source, distributed real-time search and analytics engine for the cloud based on Apache Lucene which provides full text search capabilities. It is document oriented and schema free. Asciidoctor is a pure Ruby processor for converting AsciiDoc source files and strings into HTML 5, DocBook 4.5 and other formats. Apart of Asciidoctor Ruby part, there is an Asciidoctor-java-integration project which let us call Asciidoctor functions from Java without noticing that Ruby code is being executed. In this post we are going to see how we can use Elasticsearch over AsciiDocdocuments to make them searchable by their header information or by their content. Let's add required dependencies: junit junit 4.11 test com.googlecode.lambdaj lambdaj 2.3.3 org.elasticsearch elasticsearch 0.90.1 org.asciidoctor asciidoctor-java-integration 0.1.3 Lambdaj library is used to convert AsciiDoc files to a json documents. Now we can start an Elasticsearch instance which in our case it is going to be an embedded instance. node = nodeBuilder().local(true).node(); Next step is parse AsciiDoc document header, read its content and convert them into a json document. An example of json document stored in Elasticsearch can be: { "title":"Asciidoctor Maven plugin 0.1.2 released!", "authors":[ { "author":"Jason Porter", "email":"[email protected]" } ], "version":null, "content":"= Asciidoctor Maven plugin 0.1.2 released!.....", "tags":[ "release", "plugin" ] } And for converting an AsciiDoc File to a json document we are going to useXContentBuilder class which is provided by ElasticsearchJava API to create jsondocuments programmatically. package com.lordofthejars.asciidoctor; import static org.elasticsearch.common.xcontent.XContentFactory.*; import java.io.File; import java.io.FileInputStream; import java.io.FileNotFoundException; import java.io.IOException; import java.util.List; import org.asciidoctor.Asciidoctor; import org.asciidoctor.Author; import org.asciidoctor.DocumentHeader; import org.asciidoctor.internal.IOUtils; import org.elasticsearch.common.xcontent.XContentBuilder; import ch.lambdaj.function.convert.Converter; public class AsciidoctorFileJsonConverter implements Converter { private Asciidoctor asciidoctor; public AsciidoctorFileJsonConverter() { this.asciidoctor = Asciidoctor.Factory.create(); } public XContentBuilder convert(File asciidoctor) { DocumentHeader documentHeader = this.asciidoctor.readDocumentHeader(asciidoctor); XContentBuilder jsonContent = null; try { jsonContent = jsonBuilder() .startObject() .field("title", documentHeader.getDocumentTitle()) .startArray("authors"); Author mainAuthor = documentHeader.getAuthor(); jsonContent.startObject() .field("author", mainAuthor.getFullName()) .field("email", mainAuthor.getEmail()) .endObject(); List authors = documentHeader.getAuthors(); for (Author author : authors) { jsonContent.startObject() .field("author", author.getFullName()) .field("email", author.getEmail()) .endObject(); } jsonContent.endArray() .field("version", documentHeader.getRevisionInfo().getNumber()) .field("content", readContent(asciidoctor)) .array("tags", parseTags((String)documentHeader.getAttributes().get("tags"))) .endObject(); } catch (IOException e) { throw new IllegalArgumentException(e); } return jsonContent; } private String[] parseTags(String tags) { tags = tags.substring(1, tags.length()-1); return tags.split(", "); } private String readContent(File content) throws FileNotFoundException { return IOUtils.readFull(new FileInputStream(content)); } } Basically we are building the json document by calling startObject methods to start a new object, field method to add new fields, and startArray to start an array. Then this builder will be used to render the equivalent object in json format. Notice that we are using readDocumentHeader method from Asciidoctor class which returns header attributes from AsciiDoc file without reading and rendering the whole document. And finally content field is set with all document content. And now we are ready to start indexing documents. Note that populateData method receives as parameter a Client object. This object is from Elasticsearch Java APIand represents a connection to Elasticsearch database. import static ch.lambdaj.Lambda.convert; //.... private void populateData(Client client) throws IOException { List asciidoctorFiles = new ArrayList() {{ add(new File("target/test-classes/java_release.adoc")); add(new File("target/test-classes/maven_release.adoc")); }; List jsonDocuments = convertAsciidoctorFilesToJson(asciidoctorFiles); for (int i=0; i < jsonDocuments.size(); i++) { client.prepareIndex("docs", "asciidoctor", Integer.toString(i)).setSource(jsonDocuments.get(i)).execute().actionGet(); } client.admin().indices().refresh(new RefreshRequest("docs")).actionGet(); } private List convertAsciidoctorFilesToJson(List asciidoctorFiles) { return convert(asciidoctorFiles, new AsciidoctorFileJsonConverter()); } It is important to note that the first part of the algorithm is converting all our AsciiDocfiles (in our case two) to XContentBuilder instances by using previous converter class and the method convert of Lambdaj project. If you want you can take a look to both documents used in this example in https://github.com/asciidoctor/asciidoctor.github.com/blob/develop/news/asciidoctor-java-integration-0-1-3-released.adoc and https://github.com/asciidoctor/asciidoctor.github.com/blob/develop/news/asciidoctor-maven-plugin-0-1-2-released.adoc. Next part is inserting documents inside one index. This is done by using prepareIndexmethod, which requires an index name (docs), an index type (asciidoctor), and the idof the document being inserted. Then we call setSource method which transforms theXContentBuilder object to json, and finally by calling execute().actionGet(), data is sent to database. The final step is only required because we are using an embedded instance ofElasticsearch (in production this part should not be required), which refresh the indexes by calling refresh method. After that point we can start querying Elasticsearch for retrieving information from our AsciiDoc documents. Let's start with very simple example, which returns all documents inserted: SearchResponse response = client.prepareSearch().execute().actionGet(); Next we are going to search for all documents that has been written by Alex Sotowhich in our case is one. import static org.elasticsearch.index.query.QueryBuilders.matchQuery; //.... QueryBuilder matchQuery = matchQuery("author", "Alex Soto"); QueryBuilder matchQuery = matchQuery("author", "Alexander Soto"); Note that I am searching for field author the string Alex Soto, which returns only one. The other document is written by Jason. But it is interesting to say that if you search for Alexander Soto, the same document will be returned; Elasticsearch is smart enough to know that Alex and Alexander are very similar names so it returns the document too. More queries, how about finding documents written by someone who is called Alex, but not Soto. import static org.elasticsearch.index.query.QueryBuilders.fieldQuery; //.... QueryBuilder matchQuery = fieldQuery("author", "+Alex -Soto"); And of course no results are returned in this case. See that in this case we are using afield query instead of a term query, and we use +, and - symbols to exclude and include words. Also you can find all documents which contains the word released on title. import static org.elasticsearch.index.query.QueryBuilders.matchQuery; //.... QueryBuilder matchQuery = matchQuery("title", "released"); And finally let's find all documents that talks about 0.1.2 release, in this case only one document talks about it, the other one talks about 0.1.3. QueryBuilder matchQuery = matchQuery("content", "0.1.2"); Now we only have to send the query to Elasticsearch database, which is done by using prepareSearch method. SearchResponse response = client.prepareSearch("docs") .setTypes("asciidoctor") .setQuery(matchQuery) .execute() .actionGet(); SearchHits hits = response.getHits(); for (SearchHit searchHit : hits) { System.out.println(searchHit.getSource().get("content")); } Note that in this case we are printing the AsciiDoc content through console, but you could use asciidoctor.render(String content, Options options) method to render the content into required format. So in this post we have seen how to index documents using Elasticsearch, how to get some important information from AsciiDoc files using Asciidoctor-java-integration project, and finally how to execute some queries to inserted documents. Of course there are more kind of queries in Elasticsearch, but the intend of this post wasn't to explore all possibilities of Elasticsearch. Also as corollary, note how important it is using AsciiDoc format for writing your documents. Without much effort you can build a search engine for your documentation. On the other side, imagine all code that would be required to implement the same using any proprietary binary format like Microsoft Word. So we have shown another reason to use AsciiDoc instead of other formats.

Hadoop provides a Java native API to support file system operations..

avro has a little-known gem of a feature which allows you to control which fields in an avro record are used for partitioning , sorting and grouping in mapreduce. the following figure gives a quick refresher as to what these terms mean. oh, and don’t take the placement of the “sorting” literally - sorting actually occurs on both the map and reduce side - but it’s always performed in the context of a specific partition (i.e. for a specific reducer). by default all the fields in an avro map output key are used for partitioning, sorting and grouping in mapreduce. let’s walk through an example and see how this works. you’ll begin with a simple schema github source : {"type": "record", "name": "com.alexholmes.avro.weathernoignore", "doc": "a weather reading.", "fields": [ {"name": "station", "type": "string"}, {"name": "time", "type": "long"}, {"name": "temp", "type": "int"}, {"name": "counter", "type": "int", "default": 0} ] } we’re going to see what happens when we run this code against a small sample data set, which we’ll generate using avro code github source : file input = tmpfolder.newfile("input.txt"); avrofiles.createfile(input, weathernoignore.schema$, arrays.aslist( weathernoignore.newbuilder().setstation("sfo").settime(1).settemp(3).build(), weathernoignore.newbuilder().setstation("iad").settime(1).settemp(1).build(), weathernoignore.newbuilder().setstation("sfo").settime(2).settemp(1).build(), weathernoignore.newbuilder().setstation("sfo").settime(1).settemp(2).build(), weathernoignore.newbuilder().setstation("sfo").settime(1).settemp(1).build() ).toarray()); to understand how avro is partitioning, sorting and grouping the data, we’ll write an identity mapper and reducer, with a small enhancement to the reducer to increment the counter field for each record we see in an individual reducer instance github source : package com.alexholmes.avro.sort.basic; import com.alexholmes.avro.weathernoignore; import org.apache.avro.mapred.avrokey; import org.apache.avro.mapred.avrovalue; import org.apache.avro.mapreduce.avrojob; import org.apache.avro.mapreduce.avrokeyinputformat; import org.apache.avro.mapreduce.avrokeyoutputformat; import org.apache.hadoop.fs.path; import org.apache.hadoop.io.nullwritable; import org.apache.hadoop.mapreduce.job; import org.apache.hadoop.mapreduce.mapper; import org.apache.hadoop.mapreduce.reducer; import org.apache.hadoop.mapreduce.lib.input.fileinputformat; import org.apache.hadoop.mapreduce.lib.output.fileoutputformat; import java.io.ioexception; public class avrosort { private static class sortmapper extends mapper, nullwritable, avrokey, avrovalue> { @override protected void map(avrokey key, nullwritable value, context context) throws ioexception, interruptedexception { context.write(key, new avrovalue(key.datum())); } } private static class sortreducer extends reducer, avrovalue, avrokey, nullwritable> { @override protected void reduce(avrokey key, iterable> values, context context) throws ioexception, interruptedexception { int counter = 1; for (avrovalue weathernoignore : values) { weathernoignore.datum().setcounter(counter++); context.write(new avrokey(weathernoignore.datum()), nullwritable.get()); } } } public boolean runmapreduce(final job job, path inputpath, path outputpath) throws exception { fileinputformat.setinputpaths(job, inputpath); job.setinputformatclass(avrokeyinputformat.class); avrojob.setinputkeyschema(job, weathernoignore.schema$); job.setmapperclass(sortmapper.class); avrojob.setmapoutputkeyschema(job, weathernoignore.schema$); avrojob.setmapoutputvalueschema(job, weathernoignore.schema$); job.setreducerclass(sortreducer.class); avrojob.setoutputkeyschema(job, weathernoignore.schema$); job.setoutputformatclass(avrokeyoutputformat.class); fileoutputformat.setoutputpath(job, outputpath); return job.waitforcompletion(true); } } if you look at the output of the job below, you’ll see that the output is sorted across all the fields, and that the sorting is in field ordinal order. what this means is that when mapreduce is sorting these records, it compares the station field first, then the time field second, and so on according to the ordering of the fields in the avro schema. this is pretty much what you’d expect if you write your own complex writable type, and your comparator compared all the fields in order. {"station": "iad", "time": 1, "temp": 1, "counter": 1} {"station": "sfo", "time": 1, "temp": 1, "counter": 1} {"station": "sfo", "time": 1, "temp": 2, "counter": 1} {"station": "sfo", "time": 1, "temp": 3, "counter": 1} {"station": "sfo", "time": 2, "temp": 1, "counter": 1} oh, and before we move on notice that the value for the counter field is always 1 , meaning that each reducer was only fed a single key/vaue pair, which makes sense since our identity mapper only emitted a single value for each key, the keys are unique, and the mapreduce partitioner, sorter and grouper were using all the fields in the record. excluding fields for sorting avro gives us the ability to indicate that specific fields should be ignored when performing ordering functions. in mapreduce these fields are ignored for sorting/partitioning and grouping in mapreduce, which basically means that we have the ability to perform secondary sorting. let’s examine the following schema github source : {"type": "record", "name": "com.alexholmes.avro.weather", "doc": "a weather reading.", "fields": [ {"name": "station", "type": "string"}, {"name": "time", "type": "long"}, {"name": "temp", "type": "int", "order": "ignore"}, {"name": "counter", "type": "int", "order": "ignore", "default": 0} ] } it’s pretty much identical to the first schema, the only difference being that the last two fields are flagged as being “ignored” for sorting/partitioning/grouping. let’s run the same (other than modified to work with the different schema) mapreduce code github source as above against this new schema and examine the outputs. {"station": "iad", "time": 1, "temp": 1, "counter": 1} {"station": "sfo", "time": 1, "temp": 3, "counter": 1} {"station": "sfo", "time": 1, "temp": 2, "counter": 2} {"station": "sfo", "time": 1, "temp": 1, "counter": 3} {"station": "sfo", "time": 2, "temp": 1, "counter": 1} there are a couple of notable differences between this output, and the output from the previous schema which didn’t have any ignored fields. first, it’s clear that the temp field isn’t being used in the sorting, which makes sense since we specified that it should be ignored in the schema. however, more interestingly, note the value of the counter field. all records that had identical station and time values went to the same reducer invocation, evidenced by the increasing value of counter . this is essentially secondary sort! now, all of this greatness isn’t without some limitations: you can’t support two mapreduce jobs that use the same avro key, but have different sorting/partitioning/grouping requirements. although it’s conceivable that you could create a new instance of the avro schema and set the ignored flags for these fields yourself. the partitioner, sorter and grouping functions in mapreduce all work off of the same fields (i.e. they all ignore fields that set as ignored in the schema). this means that your options for secondary sorting are limited. for example, you wouldn’t be able to partition all stations to the same reducer, and then group by station and time. ordering uses a field’s ordinal position to determine its order within the overall set of fields to be ordered. in other words, in a two-field record, the first field is always compared before the second. there’s no way to change this behavior other than flipping the order of the fields in the record. having said all of that - the “ignoring fields” feature for sorting is pretty awesome, and something that will no doubt come in handy in my future mapreduce work.

this is the first blog post in a series which looks at some data organization patterns in mapreduce. we’ll look at how to bucket output across multiple files in a single task, how to multiplex data across multiple files, and also how to coalesce data. these are all common patterns that are useful to have in your mapreduce toolkit. we’ll kick things off with a look at bucketing data outputs in your map or reduce tasks. by default when using a fileoutputformat-derived outputformat (such as textoutputformat), all the outputs for a reduce task (or a map task in a map-only job) are written to a single file in hdfs. imagine a situation where you have user activity logs being streamed into hdfs, and you want to write a mapreduce job to better organize the incoming data. as an example a large organization with multiple products may want to bucket the logs based on the product. to do this you’ll need the ability to write to multiple output files in a single task. let’s take a look at how we can make that happen. multipleoutputformat there are a few ways you can achieve your goal, and the first option we’ll look at is the multipleoutputformat class in hadoop. this is an abstract class that lets you do the following: define the output path for each and every key/value output record being emitted by a task. incorporate the input paths into the output directory for map-only jobs. redefine the key and value that are used to write to the underlying recordwriter . this is useful in situations where you want to remove data from the outputs as it duplicates data in the filename. for each output path, define the recordwriter that should be used to write the outputs. ok enough with the words - let’s look at some data and code. first up is the simple data we’ll use in our example - imagine you work at a fruit market with locations in multiple cities, and you have a purchase transaction stream which contains the store location along with the fruit that was purchased. cupertino apple sunnyvale banana cupertino pear to help bucket your data for future analysis, you want to bin each record into city-specific files. for the simple data set above you don’t want to filter, project or transform your data, just bucket it out, so a simple identity map-only job will do the job. to force more than one mapper, we’ll write the data to two separate files. $ tab="$(printf '\t')" $ hdfs -put - file1.txt << eof cupertino${tab}apple sunnyvale${tab}banana eof $ hdfs -put - file2.txt << eof cupertino${tab}pear eof here’s the code which will let you write city-specific output files. import org.apache.commons.lang.stringutils; import org.apache.hadoop.conf.configuration; import org.apache.hadoop.conf.configured; import org.apache.hadoop.fs.filesystem; import org.apache.hadoop.fs.path; import org.apache.hadoop.io.text; import org.apache.hadoop.mapred.*; import org.apache.hadoop.mapred.lib.identitymapper; import org.apache.hadoop.mapred.lib.multipletextoutputformat; import org.apache.hadoop.util.progressable; import org.apache.hadoop.util.tool; import org.apache.hadoop.util.toolrunner; import java.io.ioexception; import java.util.arrays; /** * an example of how to use {@link org.apache.hadoop.mapred.lib.multipleoutputformat}. */ public class mofexample extends configured implements tool { /** * create output files based on the output record's key name. */ static class keybasedmultipletextoutputformat extends multipletextoutputformat { @override protected string generatefilenameforkeyvalue(text key, text value, string name) { return key.tostring() + "/" + name; } } /** * the main job driver. */ public int run(final string[] args) throws exception { string csvinputs = stringutils.join(arrays.copyofrange(args, 0, args.length - 1), ","); path outputdir = new path(args[args.length - 1]); jobconf jobconf = new jobconf(super.getconf()); jobconf.setjarbyclass(mofexample.class); jobconf.setnumreducetasks(0); jobconf.setmapperclass(identitymapper.class); jobconf.setinputformat(keyvaluetextinputformat.class); jobconf.setoutputformat(keybasedmultipletextoutputformat.class); fileinputformat.setinputpaths(jobconf, csvinputs); fileoutputformat.setoutputpath(jobconf, outputdir); return jobclient.runjob(jobconf).issuccessful() ? 0 : 1; } /** * main entry point for the utility. * * @param args arguments * @throws exception when something goes wrong */ public static void main(final string[] args) throws exception { int res = toolrunner.run(new configuration(), new mofexample(), args); system.exit(res); } } run this code and you’ll see the following files in hdfs, where /output is the job output directory: $ hadoop fs -lsr /output /output/cupertino/part-00000 /output/cupertino/part-00001 /output/sunnyvale/part-00000 if you look at the output files you’ll see that the files contain the correct buckets. $ hadoop fs -lsr /output/cupertino/* cupertino apple cupertino pear $ hadoop fs -lsr /output/sunnyvale/* sunnyvale banana awesome, you have your data bucketed by store. now that we have everything working, let’s look at what we did to get there. we had to do two things to get this working: extend multipletextoutputformat this is where the magic happened - let’s look at that class again. static class keybasedmultipletextoutputformat extends multipletextoutputformat { @override protected string generatefilenameforkeyvalue(text key, text value, string name) { return key.tostring() + "/" + name; } } you are working with text, which is why you extended multipletextoutputformat , a class that in turn extends multipleoutputformat . multipletextoutputformat is a simple class which instructs the multipleoutputformat to use textoutputformat as the underlying output format for writing out the records. if you were to use multipleoutputformat as-is it behaves as if you were using the regular textoutputformat , which is to say that it’ll only write to a single output file. to write data to multiple files you had to extend it, as with the example above. the generatefilenameforkeyvalue method allows you to return the output path for an input record. the third argument, name , is the original fileoutputformat -created filename, which is in the form “part-nnnnn”, where “nnnnn” is the task index, to ensure uniqueness. to avoid file collisions, it’s a good idea to make sure your generated output paths are unique, and leveraging the original output file is certainly a good way of doing this. in our example we’re using the key as the directory name, and then writing to the original fileoutputformat filename within that directory. specify the outputformat the next step was easy - specify that this output format should be used for your job: jobconf.setoutputformat(keybasedmultipletextoutputformat.class); earlier we also mentioned that you can use the input path as part of the output path, which we will look at next. using the input filename as part of the output filename in map-only jobs what if we wanted to keep the input filename as part of the output filename? this only works for map-only jobs, and can be accomplished by overriding the getinputfilebasedoutputfilename method. let’s look at the following code to understand how this method fits into the overall sequence of actions that the multipleoutputformat class performs: public void write(k key, v value) throws ioexception { // get the file name based on the key string keybasedpath = generatefilenameforkeyvalue(key, value, myname); // get the file name based on the input file name string finalpath = getinputfilebasedoutputfilename(myjob, keybasedpath); // get the actual key k actualkey = generateactualkey(key, value); v actualvalue = generateactualvalue(key, value); recordwriter rw = this.recordwriters.get(finalpath); if (rw == null) { // if we don't have the record writer yet for the final path, create // one // and add it to the cache rw = getbaserecordwriter(myfs, myjob, finalpath, myprogressable); this.recordwriters.put(finalpath, rw); } rw.write(actualkey, actualvalue); }; the getinputfilebasedoutputfilename method is called with the output of generatefilenameforkeyvalue , which contains our already-customized output file. our new keybasedmultipletextoutputformat can now be updated to override getinputfilebasedoutputfilename and append the original input filename to the output filename: static class keybasedmultipletextoutputformat extends multipletextoutputformat { @override protected string generatefilenameforkeyvalue(object key, object value, string name) { return key.tostring() + "/" + name; } @override protected string getinputfilebasedoutputfilename(jobconf job, string name) { string infilename = new path(job.get("map.input.file")).getname(); return name + "-" + infilename; } if you run with your modified outputformat class you’ll see the following files in hdfs, confirming that the input filenames are now concatenated to the end of each output file. $ hadoop fs -lsr /output /output/cupertino/part-00000-file1.txt /output/cupertino/part-00001-file2.txt /output/sunnyvale/part-00000-file1.txt the implementation of getinputfilebasedoutputfilename in multipleoutputformat doesn’t do anything interesting by default, but if you set the value of the mapred.outputformat.numoftrailinglegs configurable to an integer greater than 0, then the getinputfilebasedoutputfilename will use part of the input path as the output path. let’s see what happens when we set the value to 1: jobconf.setint("mapred.outputformat.numoftrailinglegs", 1); the output files in hdfs now exactly mirror the input files used for the job: $ hadoop fs -lsr /output /output/file1.txt /output/file2.txt if we set mapred.outputformat.numoftrailinglegs to 2, and our input files exist in the /inputs directory, then our output directory looks like this: $ hadoop fs -lsr /output /output/input/file1.txt /output/input/file2.txt basically as you keep incrementing mapred.outputformat.numoftrailinglegs , then multipleoutputformat will continue to go up the parent directories of the input file and use them in the output path. modifying the output key and value it’s very possible that the actual key and value you want to emit are different from those that were used to determine the output file. in our example, we took the output key and wrote to a directory using the key name. if you do that keeping the key in the output file may be redundant. how would we modify the output record so that the key isn’t written? multipleoutputformat has your back with the generateactualkey method. class keybasedmultipletextoutputformat extends multipletextoutputformat { @override protected string generatefilenameforkeyvalue(text key, text value, string name) { return key.tostring() + "/" + name; } @override protected text generateactualkey(text key, text value) { return null; } } the returned value from this method replaces the key that’s supplied to the underlying recordwriter , so if you return null as in the above example, no key will be written to the file. $ hadoop fs -lsr /output/cupertino/* apple pear $ hadoop fs -lsr /output/sunnyvale/* banana you can achieve the same result for the output value by overriding the generateactualvalue method. changing the recordwriter in our final step we’ll look at how you can leverage multiple recordwriter classes for different output files. this is accomplished by overriding the getrecordwriter method. in the example below we’re leveraging the same textoutputformat for all the files, but it gives you a sense of what can be accomplished. static class keybasedmultipletextoutputformat extends multipletextoutputformat { @override protected string generatefilenameforkeyvalue(text key, text value, string name) { return key.tostring() + "/" + name; } @override public recordwriter getrecordwriter(filesystem fs, jobconf job, string name, progressable prog) throws ioexception { if (name.startswith("apple")) { return new textoutputformat().getrecordwriter(fs, job, name, prog); } else if (name.startswith("banana")) { return new textoutputformat().getrecordwriter(fs, job, name, prog); } return super.getrecordwriter(fs, job, name, prog); } } conclusion when using multipleoutputformat , give some thought to the number of distinct files that each reducer will create. it would be prudent to plan your bucketing so that you have a relatively small number of files. in this post we extended multipletextoutputformat , which is a simple extension of multipleoutputformat that supports text outputs. multiplesequencefileoutputformat also exists to support sequencefiles in a similar fashion. so what are the shortcomings with the multipleoutputformat class? if you have a job that uses both map and reduce phases, then multipleoutputformat can’t be used in the map-side to write outputs. of course, multipleoutputformat works fine in map-only jobs. all recordwriter classes must support exactly the same output record types. for example, you wouldn’t be able to support a recordwriter that emitted for one output file, and have another recordwriter that emitted . multipleoutputformat exists in the mapred package, so it won’t work with a job that requires use of the mapreduce package. all is not lost if you bump into either one of these issues, as you’ll discover in the next blog post.

Hebrew search is not an easy task, and HebMorph is a project I started several years ago to address that problem. After a certain period of inactivity I'm back actively working on it. I'm also happy to say there are already several live systems using it to enable Hebrew searches in their applications. This post is a short step-by-step guide on how to use HebMorph in an ElasticSearch installation. There are quite a few configuration options and things to consider when enabling Hebrew search, most are in the realm of performance vs relevance trade-offs, but I'll talk about those in a separate post. 0. What exactly is HebMorph HebMorph is a project a bit wider than just providing a Hebrew search plugin for ElasticSearch, but for the purpose of this post let us treat it in that narrow aspect. HebMorph has 3 main parts - the hspell dictionary files, the hebmorph-core package which is a wrapper around the dictionary files with important bits that allow for locating words even if they weren't written exactly as they appear in the dictionary, and the hebmorph-lucene package which contains various tools for processing streams of text into Lucene tokens - the searchable parts. To enable Hebrew search from ElasticSearch we are going to need to use the Hebrew analyzer class HebMorph provides to analyze incoming Hebrew texts. That is done by providing ElasticSearch with the HebMorph packages and then telling it to use the Hebrew analyzer on text fields as needed. 1. Get HebMorph and hspell At the moment you will have to compile HebMorph from sources yourself using Maven. In the future we might upload it to a centralized repository, but since we still actively working on a lot of stuff there it is still a bit too early for that. Probably the easiest way to get HebMorph is to do git clone from the main repository. The repository is located at https://github.com/synhershko/HebMorph and includes the latest hspell files already under /hspell-data-files. If you are new to git GitHub offers great tutorials for getting started with it, and they also enable you to download the entire source tree as a zip or a tarball. Once you have the sources, run mvn package or mvn install to create 2 jars - hebmorph-core and hebmorph-lucene. Those 2 packages are required before moving on to the next step. 2. Create an ElasticSearch plugin In this step we will create a new plugin which we will use in the next step to create the Hebrew analyzers in. If you already have a plugin you wish to use, skip to the next step. ElasticSearch plugins are compiled Java packages you simply drop to the plugins folder of your ElasticSearch installation and it gets detected automatically by the ElasticSearch instance once it is initialized. If you are new to this, you might want to read up a bit on that in the official ElasticSearch documentation. Here is a great guide to start with: http://jfarrell.github.io/ The gist of this is having a Java project with a es-plugin.properties file embedded as a resource and pointing to class that tells ElasticSearch what classes to load as plugins, and their plugin type. In the next section we will use this to add our own Analyzer implementation which makes use of HebMorph's capabilities. 3. Creating an Hebrew Analyzer HebMorph already comes with MorphAnalyzer - an Analyzer implementation which takes care of Hebrew-aware tokenization, lemmatization and whatnot. Because it is highly configurable, personally I prefer re-implementing it in the ElasticSearch plugin so it is easier to change the configurations in code. In case you wondered, I'm not planning in supporting external configurations for this as it is too subtle and you should really know what you are doing there. Don't forget to add dependencies to hebmorph-core and hebmorph-lucene to your project. My common Analyzer setup for Hebrew search looks like this: public abstract class HebrewAnalyzer extends ReusableAnalyzerBase { protected enum AnalyzerType { INDEXING, QUERY, EXACT } private static final DictRadix prefixesTree = LingInfo.buildPrefixTree(false); private static DictRadix dictRadix; private final StreamLemmatizer lemmatizer; private final LemmaFilterBase lemmaFilter; protected final Version matchVersion; protected final AnalyzerType analyzerType; protected final char originalTermSuffix = '$'; static { try { dictRadix = Loader.loadDictionaryFromHSpellData(new File(resourcesPath + "hspell-data-files"), true); } catch (IOException e) { // TODO log } } protected HebrewAnalyzer(final AnalyzerType analyzerType) throws IOException { this.matchVersion = matchVersion; this.analyzerType = analyzerType; lemmatizer = new StreamLemmatizer(null, dictRadix, prefixesTree, null); lemmaFilter = new BasicLemmaFilter(); } @Override protected TokenStreamComponents createComponents(final String fieldName, final Reader reader) { // on query - if marked as keyword don't keep origin, else only lemmatized (don't suffix) // if word termintates with $ will output word$, else will output all lemmas or word$ if OOV if (analyzerType == AnalyzerType.QUERY) { final StreamLemmasFilter src = new StreamLemmasFilter(reader, lemmatizer, null, lemmaFilter); src.setAlwaysSaveMarkedOriginal(true); src.setSuffixForExactMatch(originalTermSuffix); TokenStream tok = new SuffixKeywordFilter(src, '$'); return new TokenStreamComponents(src, tok); } if (analyzerType == AnalyzerType.EXACT) { // on exact - we don't care about suffixes at all, we always output original word with suffix only final HebrewTokenizer src = new HebrewTokenizer(reader, prefixesTree, null); TokenStream tok = new NiqqudFilter(src); tok = new LowerCaseFilter(matchVersion, tok); tok = new AlwaysAddSuffixFilter(tok, '$', false); return new TokenStreamComponents(src, tok); } // on indexing we should always keep both the stem and marked original word // will ignore $ && will always output all lemmas + origin word$ // basically, if analyzerType == AnalyzerType.INDEXING) final StreamLemmasFilter src = new StreamLemmasFilter(reader, lemmatizer, null, lemmaFilter); src.setAlwaysSaveMarkedOriginal(true); TokenStream tok = new SuffixKeywordFilter(src, '$'); return new TokenStreamComponents(src, tok); } public static class HebrewIndexingAnalyzer extends HebrewAnalyzer { public HebrewIndexingAnalyzer() throws IOException { super(AnalyzerType.INDEXING); } } public static class HebrewQueryAnalyzer extends HebrewAnalyzer { public HebrewQueryAnalyzer() throws IOException { super(AnalyzerType.QUERY); } } public static class HebrewExactAnalyzer extends HebrewAnalyzer { public HebrewExactAnalyzer() throws IOException { super(AnalyzerType.EXACT); } } } You may notice how I created 3 separate analyzers - one for indexing, one for querying and the last for exact querying. I'll be talking more about this in future posts, but the idea is to be able to provide flexibility on querying while still allow for correct indexing. Configuring the analyzers to be picked up from ElasticSearch is rather easy now. First, you need to wrap each analyzer in a "provider", like so: public class HebrewQueryAnalyzerProvider extends AbstractIndexAnalyzerProvider { private final HebrewAnalyzer.HebrewQueryAnalyzer hebrewAnalyzer; @Inject public HebrewQueryAnalyzerProvider(Index index, @IndexSettings Settings indexSettings, Environment env, @Assisted String name, @Assisted Settings settings) throws IOException { super(index, indexSettings, name, settings); hebrewAnalyzer = new HebrewAnalyzer.HebrewQueryAnalyzer(); } @Override public HebrewAnalyzer.HebrewQueryAnalyzer get() { return hebrewAnalyzer; } } After you've created such providers for all types of analyzers, create an AnalysisBinderProcessor like this (or update your existing one with definitions for the Hebrew analyzers): public class MyAnalysisBinderProcessor extends AnalysisModule.AnalysisBinderProcessor { private final static HashMap> languageAnalyzers = new HashMap<>(); static { languageAnalyzers.put("hebrew", HebrewIndexingAnalyzerProvider.class); languageAnalyzers.put("hebrew_query", HebrewQueryAnalyzerProvider.class); languageAnalyzers.put("hebrew_exact", HebrewExactAnalyzerProvider.class); } public static boolean analyzerExists(final String analyzerName) { return languageAnalyzers.containsKey(analyzerName); } @Override public void processAnalyzers(final AnalyzersBindings analyzersBindings) { for (Map.Entry> entry : languageAnalyzers.entrySet()) { analyzersBindings.processAnalyzer(entry.getKey(), entry.getValue()); } } } Don't forget to update your Plugin class to catch the AnalysisBinderProcessor - it should look something like this (plus any other stuff you want to add there): public class MyPlugin extends AbstractPlugin { @Override public String name() { return "my-plugin"; } @Override public String description() { return "Implements custom actions required by me"; } @Override public void processModule(Module module) { if (module instanceof AnalysisModule) { ((AnalysisModule)module).addProcessor(new MyAnalysisBinderProcessor()); } } } 4. Using the Hebrew analyzers Compile the ElasticSearch plugin and drop it along with its dependencies in a folder under the /plugins folder of ElasticSearch. You now have 3 new types of analyzers at your disposal: "hebrew", "hebrew_query" and "hebrew_exact". For indexing, you want to use the "hebrew" analyzer. In your mapping, you can define a certain field or an entire set of fields to use that specific analyzer by setting the analyzer for that field. You can also leave the analyzer configuration blank, and specify the analyzer to use for those fields with unspecified analyzer using the _analyzer field in the index request. See more about both here and here. The "hebrew" analyzer will expand each term to all recognized lemmas; in case the word wasn't recognized it will try to tolerate spelling errors or missing Yud/Vav - most of the time it will be successful (with some rate of false positives, which the lemma-filters should remove to some degree). Some words will still remain unrecognized and thus will be indexed as-is. When querying using a QueryString query you can specify what analyzer to use - use the "hebrew_query" or "hebrew_exact" analyzer. The former will perform lemma expansion similar to the indexing analyzer, and the latter will avoid that and allow you to perform exact matches (useful when searching for names or exact phrases). I pretty much ignored a lot of the complexity involved in fine tuning searches for Hebrew, and many very cool things HebMorph allows you to do with Hebrew search for the sake of focus. I will revisit them in a later blog post. 5. Administration The hspell dictionary files are looked up by a physical location on disk - you will need to provide a path they are saved at. Since dictionaries update, it is sometimes easier to update them that way in a distributed environment like the one I'm working with. It may be desirable to have them compiled within the same jar file as the code itself - I'll be happy to accept a pull request to do that. The code above is working with ElasticSearch 0.90 GA and Lucene 4.2.1. I also had it running on earlier versions of both technologies, but may had to make a few minor changes. I assume the samples would break on future versions and I'll probably don't have much time going back and keeping it up to date, but bear in mind most of the time the changes are minor and easy to understand and make by yourself. Both HebMorph and the hspell dictionary are released under the AGPL3. For any questions on licensing, feel free to contact me.

I just spent some time putting together some basic Java code to read some data from HDFS. Pretty basic stuff. No map reduce involved. Pretty boilerplate code like the stuff from this popular tutorial on the topic. No matter what, I kept hitting my head on this error: Exception in thread “main” java.lang.IllegalArgumentException: Wrong FS: hdfs://localhost:9000/user/hadoop/DOUG_SVD/out.txt, expected: file:/// If you checkout the tutorial above, what’s supposed to be happening is that an instance of Hadoop’s Configuration should encounter a fs.default.name property, in one of the config files its given. The Configuration should realize that this property has a value of hdfs://localhost:9000. When you use the Configuration to create a Hadoop FileSystem instance, it should happily read this property from Configuration and process paths from HDFS. That’s a long way of saying these three lines of Java code: // pickup config files off classpath Configuration conf = new Configuration() // explicitely add other config files conf.addResource("/home/hadoop/conf/core-site.xml"); // create a FileSystem object needed to load file resources FileSystem fs = FileSystem.get(conf); // load files and stuff below! Well… My Hadoop config files (core-site.xml) appear setup correctly. It appears to be in my CLASSPATH. I’m even trying to explicitly add the resource. Basically I’ve followed all the troubleshooting tips you’re supposed to follow when you encounter this exception. But I’m STILL getting this exception. Head meet wall. This has to be something stupid. Troubleshooting Hadoop’s Configuration & FileSystem Objects Well before I reveal my dumb mistake in the above code, it turns out there’s some helpful functions to help debug these kind of problems: As Configuration is just a bunch of key/value pairs from a set of resources, its useful to know what resources it thinks it loaded and what properties it thinks it loaded from those files. getRaw() — return the raw value for a configuration item (like conf.getRaw("fs.default.name")) toString() — Configuration‘s toString shows the resources loaded You can similarly checkout FileSystem‘s helpful toString method. It nicely lays out where it thinks its pointing (native vs HDFS vs S3 etc). So if you similarly are looking for a stupid mistake like I was, pepper your code with printouts of these bits of info. They will at least point you in a new direction to search for your dumb mistake. Drumroll Please Turns out I missed the crucial step of passing a Path object not a String to addResource. They appear to do slightly different things. Adding a String adds a resource relative to the classpath. Adding a Path is used to add a resource at an absolute location and does not consider the classpath. So to explicitly load the correct config file, the code above gets turned into (drumroll please): // pickup config files off classpath Configuration conf = new Configuration() // explicitely add other config files // PASS A PATH NOT A STRING! conf.addResource(new Path("/home/hadoop/conf/core-site.xml")); FileSystem fs = FileSystem.get(conf); // load files and stuff below! Then Tada! everything magically works! Hopefully these tips can save you the next time you encounter these kinds of problems.

Introduction The IT buzzword of 2012 is without a doubt Big Data. It’s new and here to stay, and for a good reason. Big data is data that exceeds the processing capacity of conventional database systems. Great examples are CERN with the Large Hadron Collider, whose experiments generate 25 petabytes of data annually, or Walmart, which handles more than one million customer transaction every hour. Problems These vast amounts of data leave us with two problems. Problem 1: To gain value from this data, one must choose an alternative way to process it. The value of big data to an organization falls into two categories: analytical use, and enabling new products. Big data analytics can reveal insights hidden previously by data too costly to process, such as peer influence among customers, revealed by analyzing shoppers’ transactions, social and geographical data. Being able to process every item of data in reasonable time removes the troublesome need for sampling and promotes an investigative approach to data, in contrast to the somewhat static nature of running predetermined reports. Problem 2: The data is too big, moves too fast, or doesn’t fit the strictures of your database architectures. Remember the CERN case where the LHC produces over 25 Petabytes of data annually? No “classic” database architecture or setup is capable of holding these amounts of data. Solutions Fortunately, both problems can be solved by implementing the correct infrastructure and rethinking data storage. There are two critical factors in Big Data environments: size and speed. We already discussed the vast amounts of data and desire to be able to access and process the data fast. The latter is the main differentiator from more traditional data warehouses. Just imagine what you can do when you can access all your data real-time. Enter big data. A common Big Data implementation is an in-memory data grid that lives in a distributed cluster, ensuring both speed, by storing data in-memory, and capacity by using scalability features provided by a cluster. As a bonus, availability is ensured by using a distributed cluster. As for the data storage, there are typically two kinds: in-memory databases and in-memory data grids. But first some background. It is not a new attempt to use main memory as a storage area instead of a disk. In our daily lives there are numerous examples of main memory databases (MMDB), as they perform much faster than disk-based databases. An every day example is a mobile phone. When you SMS or call someone most mobile service providers use MMDB to get the information on your contact as soon as possible. The same applies to your phone. When someone calls you, the caller details are looked up in the contacts application, usually providing a name and sometimes a picture. In memory data grids In Memory Data Grid (IMDG) is the same as MMDB in that it stores data in main memory, but it has a totally different architecture. The features of IMDG can be summarized as follows: Data is distributed and stored on multiple servers. Each server operates in the active mode. A data model is usually object-oriented (serialized) and non-relational. According to the necessity, you often need to add or reduce servers. No traditional database features such as tables. In other words, IMDG is designed to store data in main memory, ensure scalability and store an object itself. These days, there are many IMDG products, both commercial and open source. Some of the most commonly used products are: Hazelcast (http://www.hazelcast.com) JBoss Infinispan (http://www.jboss.org/infinispan) GridGain DataGrid (http://www.gridgain.com/features/in-memory-data-grid/) VMware Gemfire (http://www.vmware.com/nl/products/application-platform/vfabric-gemfire/overview.html) Oracle Coherence (http://www.oracle.com/technetwork/middleware/coherence/overview/index.html) Gigaspaces XAP (http://www.gigaspaces.com/datagrid) Terracotta Enterprise Suite (http://terracotta.org/products/enterprise-suite) Why Memory? The main reasons for using main memory for data storage are once again the two main themes of Big Data: speed and capacity. The processing performance of main memory is 800 times faster than an HDD and up to 40 times faster than an. Moreover, the latest x86 server supports main memory of hundreds of GB per server. It is said that the limit of a traditional processing database’s (OLTP) data capacity is approximately 1 TB and that the OLTP processing data capacity would not increment well. If servers using main memory of 1 TB or larger become more commonly used, you will be able to conduct operations with the entire data placed in main memory, at least in the field of OLTP. IMDG Architecture To use main memory as a storage area, two weak points should be overcome: Limited capacity: involves data that exceeds the maximum capacity of the main memory of the server Reliability: involves data loss in case of a (system) failure. IMDG overcomes the limit of capacity by ensuring horizontal scalability using a distributed architecture, and resolves the issue of reliability through a replication system as part of the grid (or a distributed cluster). Now let’s discuss how an IMDG actually works. First of all, it is important to understand that an IMDG is not the same as an in-memory database, also referred to as MMDB (main memory databases). Typical examples of MMDBs are Oracle TimesTen or Sap Hana. MMDBs are full database products that simply reside in memory. As a result of being a full-blown database, they also carry the weight and overhead of database management features. IMDG is different. No tables, indexes, triggers, stored procedures, process managers etc. Just plain storage. The data model used in IMDG is key-value pairs. A key-value pair is a list with only two parts: a key and a value. The key can be used for storing and retrieving the values in the list. A key can be compared to the index or primary key of a table in a database. Note that IMDG are closely tied to development environments such as Java as the key-value pairs are represented by the structures provided by such a programming environment. Most IMDGs are written in Java, and can only be used within other Java applications. Therefore, the values of key-value pairs can be anything supported by Java, ranging from simple data types such as a string or number, to complex objects. This overcomes the two important hurdles: as you can store complex Java objects as value, there’s no need to translate these objects into a relational datamodel (which is the case in more traditional applications using a database for storage). Furthermore, the seeming limitation of being able to store only one value per key, is actually no limitation at all. Large memory sizes Most of the products introduced above use Java as an implementation language. Java reserves and uses a part of the RAM (internal memory) for dynamic memory allocation. This reserved memory space is called the Java heap. All runtime objects created by a Java application are stored in heap. Using large amounts of data causes two problems. Size limitation: By default, the heap size is 128 MB, but for current business applications, this limit is reached easily. Once the heap is “full”, no new objects can be created and the Java application will show some nasty errors. Performance: It is possible to increase the size of the heap, but this introduces some new problems. When a heap reaches a size of more than 4 gigabytes, Java will have serious issues with memory managements, causing your application to slow down or even freeze. Java has a feature called Garbage Collector, which periodically scans the heap and checks each object if it is still valid and being used. If not, the garbage collector removes the object and defragments the newly available space. The problem is, the larger the heap size, the more work to do for the garbage collector, resulting in performance degradation. Imagine a large bank has a Java application that manages customers, accounts and transactions. We have seen that an IMDG allows the application to store and access all data very quickly by caching it in memory, instead of storing the data in relatively slow databases. Let’s assume the combined data has a size of 40 gigabytes. Storing it in heap is simply not possible, considering the performance penalties of Java’s memory management capabilities. The graph below illustrates the garbage collection pause time when placing cached data in heap: Terracotta’s BigMemory product has a method to overcome these limitations. The method is to use an off-heap memory (direct buffer). Data will not be stored in Java’s heap, but directly in the available internal memory (RAM). Since, this is not subject to Java’s garbage collector, there are no performance penalties. The differences on performance are significant, as can be seen in the graph below: Using off-heap storage has some major benefits: You can use all the available memory on your machine, not just the memory that is allocated to the heap (usually less that 512 Mb). This allows you to store more data in a in-memory data grid, greatly speeding up your application. The heap can be relieved by storing data in native memory, speeding up Java applications as less heap space has to be garbage collected. Clustering, fail over and high availability So far, we have seen IMDG features that are applicable to a single server. However, the real power of IMDG lies in it’s networking and clustering capabilities, providing features as data replication, data synchronization between clients, fail over and high availability. To achieve this, a cluster of servers (or server array) acts a backbone of the infrastructure. Applications (that still can have their own IMDG or off-heap cache) that are connected to the cluster can share, replicate and backup their data with either the cluster or other applications. The graph below depicts a typical setup using Terracotta's BigMemory: The caches on the application servers are usually referred to as “level 1” cache, while the data cache on the server array is referred to as “level 2” cache. There are many different scenarios possible for storing, clustering, synchronizing and replicating data. Covering all these topics goes far beyond the scope of this article. For more information, consult the technical documentation of the product of your choice. Conclusion Big Data brings us some new challenges. First of all, storing and accessing vast amounts of data makes us rethink traditional methods and technologies. Next, there’s the question what to do with all the available data. The potential value for marketing, financial and other businesses is huge. In order to facilitate Big Data, in-memory data grids are considered the best option. IMDGs with off-heap storage are even more powerful, allowing data centric enterprise application to overcome certain limits of the Java platform, such as memory and performance constraints. As the amount of data that (large) companies produce and store, grows exponentially, databases will hit a limit. Accessing your data without a performance penalty simply will not be possible. The answer to this is using an IMDG.

In this part, I will be briefing about different types of Database Normalizations using a sample data model. What is Database Normalization? Normalization is the process of efficiently organizing data in the database. Primary Goal of Normalization? Eliminating redundant data & ensuring meaningful data dependencies. Types of Normalization The following are the three most common normal forms in the database normalization process First Normal Form (1NF) Second Normal Form (2NF) Third Normal Form (3NF) Sample Data Model for Demonstration The following data model will be used to demonstrate all the three normal forms First Normal Form (1NF) First Normal Form (1NF) sets the very basic rules for an organized database: Create separate set of tables for each group of related data and identify each row with a unique columns [primary key] or set of columns [composite key] Eliminate duplicate columns from the table The following data model depicts the tables after 1NF rules are applied - Second Normal Form (2NF) Second Normal Form (2NF) further addresses the concept of removing duplicate data: Meet all the requirements of the first normal form Remove subsets of data that apply to multiple rows of a table and place them in separate tables Create relationships between these new tables and their predecessors through the use of foreign keys So basically the objective of the Second Normal Form is to take that is only partly dependent on the primary key and enter that data into another table. The following data model depicts the tables after 2NF rules are applied. Data from EMPLOYEE_TABLE is split into 2 tables – EMPLOYEE_TABLE and EMPLOYEE_HR_TABLE. Similarly data from CUSTOMER_TABLE is moved to CUSTOMER_TABLE and CUSTOMER_ORDER table Third Normal Form (3NF) Third normal form (3NF) goes one large step further: Meet all the requirements of the second normal form. Remove columns that are not dependent upon the primary key. The following data model depicts the tables after 3NF rules are applied. Further state and country details are moved to their own tables because they are not dependent on the primary key. Advantages of Normalizing the Database There are several advantages of normalization - Data can be stored as small atomic pieces Saves space Increases speed Reduces data anomalies Easy maintenance Other parts of this series include: Part 1 – ACID Properties Part 2 – Keys Part 4 – Database Transactions [coming soon] Part 5 – Indexes [coming soon]

When working with MapReduce one of the challenges that is encountered early-on is determining how to make your third-part JAR’s available to the map and reduce tasks. One common approach is to create a fat jar, which is a JAR that contains your classes as well as your third-party classes (see this Cloudera blog post for more details). A more elegant solution is to take advantage of the libjars option in the hadoop jar command, also mentioned in the Cloudera post at a high level. Here I’ll go into detail on the three steps required to make this work. Add libjars to the options It can be confusing to know exactly where to put libjars when running the hadoop jar command. The following example shows the correct position of this option: $ export LIBJARS=/path/jar1,/path/jar2 $ hadoop jar my-example.jar com.example.MyTool -libjars ${LIBJARS} -mytoolopt value It’s worth noting in the above example that the JAR’s supplied as the value of the libjar option are comma-separated, and not separated by your O.S. path delimiter (which is how a Java classpath is delimited). You may think that you’re done, but often times this step alone may not be enough - read on for more details! Make sure your code is using GenericOptionsParser The Java class that’s being supplied to the hadoop jar command should use the GenericOptionsParser class to parse the options being supplied on the CLI. The easiest way to do that is demonstrated with the following code, which leverages the ToolRunner class to parse-out the options: public static void main(final String[] args) throws Exception { Configuration conf = new Configuration(); int res = ToolRunner.run(conf, new com.example.MyTool(), args); System.exit(res); } t is crucial that the configuration object being passed into the ToolRunner.run method is the same one that you’re using when setting-up your job. To guarantee this, your class should use the getConf() method defined in Configurable (and implemented in Configured) to access the configuration: public class SmallFilesMapReduce extends Configured implements Tool { public final int run(final String[] args) throws Exception { Job job = new Job(super.getConf()); ... job.waitForCompletion(true); return ...; } f you don’t leverage the Configuration object supplied to the ToolRunner.run method in your MapReduce driver code, then your job won’t be correctly configured and your third-party JAR’s won’t be copied to the Distributed Cache or loaded in the remote task JVM’s. It’s the ToolRunner.run method (actually it delegates the command parsing to GenericOptionsParser) which actually parses-out the libjars argument, and adds to the Configuration object a value for the tmpjarproperty. So a quick way to make sure that this step is working is to look at the job file for your MapReduce job (there’s a link when viewing the job details from the JobTracker), and make sure that the tmpjar configuration name exists with a value identical to the path that you specified in your command. You can also use the command-line to search for the libjars configuration in HDFS $ hadoop fs -cat /_logs/history/*.xml | grep tmpjars Use HADOOP_CLASSPATH to make your third-party JAR’s available on the client-side So far the first two steps tackled what you needed to do to to make your third-party JAR’s available to the remote map and reduce task JVM’s. But what hasn’t been covered so far is making these same JAR’s available to the client JVM, which is the JVM that’s created when you run the hadoop jar command. For this to happen, you should set the HADOOP_CLASSPATH environment variable to contain the O.S. path-delimited list of third-party JAR’s. Let’s extend the commands in the first step above with the addition of setting the HADOOP_CLASSPATH environment variable: $ export LIBJARS=/path/jar1,/path/jar2 $ export HADOOP_CLASSPATH=/path/jar1:/path/jar2 $ hadoop jar my-example.jar com.example.MyTool -libjars ${LIBJARS} -mytoolopt value Note that value for HADOOP_CLASSPATH uses a Unix path delimiter of :, so modify accordingly for your platform. And if you don’t like the copy-paste above you could modify that line to substitute the commas for semi-colons: $ export HADOOP_CLASSPATH=`echo ${LIBJARS} | sed s/,/:/g`

once upon a time there was a munich pizza baker who developed a technique to beam pizza out of bright sunshine. he can produce more than a thousand pizzas per second and needs a channel to sell this amount of pizza and decides to build an online shop. mario’s initial idea is to sell pizzas, but now he is thinking about introduction of new product lines like beverages, salads and pasta. before we take a look to the validation of mario´s idea, lets take a short look at the existing online shop. mario’s online shop is based on mongodb , apache wicket and spring . mongodb is a document-oriented nosql-database . mongodb stores records not in tables as a relational database but in bson documents, which is a binary version of json (java script object notation) and very similar to the object structure in mario’s application. the usage of mongodb makes his development easier and deployment faster. the figure shows a json document which is very similar to a java object: a json document property with the according value corresponds to the java object property with the appropriate value. you can add or remove properties in your java object and this will automatically change your database schema. so there is no need to put your java object model into a relational schema via hibernate. mario also decided to build his online shop only with open-source technologies like apache wicket and spring. wicket is a very common lightweight component-based web application framework and it is closely patterned after stateful gui frameworks such as javafx . the spring framework is an open source application framework and inversion of control container for the java platform and does not impose any specific programming model. spring has become popular in the java community as an alternative to, replacement for, or even addition to the enterprise javabean (ejb) model. because of this architecture mario is able to deploy its application in a lightweight application server like tomcat or jetty . this figure shows the system landscape of mario. mario has two major system on the lefthand site there is his online shop and on the righthand site there is ‘pas’ a famous billing system. in the middle is hadoop that connects both systems together. in the business world an application normally does not stand alone. in most cases an application must communicate with others. the lean architecture of marios online shop enables him to connect the billing system ‘pas’ to his online shop. spring for apache hadoop provides this integration between the two systems online shop and ‘pas’. hadoop supports data-intensive distributed applications and implements a computational paradigm named mapreduce, where the computation is divided into many small fragments, each of them may be executed or re-executed on any node in the cluster of commodity hardware. mario uses hadoop as an etl layer that enables him to transfer gigabytes of order information into the billing system. in this case hadoop makes it possible for a financial controller to verify if all orders were billed correctly. in addition to the online shop feature mario has a real-time sales dashboard that enables him to track his sales in real time. the dashboard displays daily and monthly sales statistics for each pizza and contains a map with the geographical overview of customer activity and competitor locations. here is a walkthrough of the shop : now lets talk about mario’s incredible new idea : mario wants to sell even more pizza! and other products as well. mario decides to use lean startup methods in order to test the possible introduction of new product lines and plans an experiment to validate his new idea using a scientific approach and pure facts instead of hunches. mario´s core assumption is that customers wants to buy other products than pizza – drinks, salads and pasta. furthermore he is worried about pricing. mario contacts all customers to complete a survey and provides an incentive for the participation, a free pizza to every customer who responds to the survey. the result of the survey validated mario’s assumption – customers want to buy beverages, salads and pasta. but he also found out that his customers are willing to pay higher prices for high-quality products and that they simply love his easy shopping flow. currently a pizza order can be completed with three clicks only, so there is new riskiest assumption to validate: will a more complex shopping flow affect his sales? the figures shows a validation board. a validation board is a deceptively simple tool for testing out product ideas. furthermore a validation board tracks pivots which follows from customer feedback. mario decides to introduce beverages, salads and pasta product lines and thinks about a possibility, how he can handle the extension of the product line without destroying the easy shopping flow. that’s why mario thinks a recommendation engine is the right way for him. panels for recommendations can be integrated in the online shop without changing the shopping flow. mario hired a statistician to help him implement a recommender system for his online shop for better cross-selling. he also defined new measurement points to validate his new idea . therefore he tracks the conversion rate of orders as well as cross-selling rates and every event in the online shop is already tracked in realtime. so mario can very easily perform further experiments in order to verify more assumptions. follow the blog to see how the story continues or come to mongodb usergroup meetup in munich , february 20, 2013 or mongodb days in berlin , february 26, 2013 to get a live presentation. our talk sheds light on how to build an online recommendation engine based on mongodb and apache mahout. we’ll show which recommenders must be built to reach mario’s goal and how these can be integrated in mario’s shop infrastructure.

imagine you have a csv file that you want to manipulate. here’s a sample file we can play with: lopez,charlie,2002,11,21 parker,ward,1995,04,08 henderson,russell,2007,10,01 our goal is to transform this into the following form by combining the last three columns: lopez,charlie,20021121 parker,ward,19950408 henderson,russell,20071001 in linux this would take all of two seconds (excuse the awkward awk command): shell$ awk -f"," '{ print $1","$2","$3$4$5 }' people.txt what if you wanted to quickly do the same in hdfs - and let’s assume you want to write the results back to hdfs. one approach would be to use the hdfs cli to stream the inputs into awk, and stream the awk output back into hdfs. you could do this with the hdfs cat and put - options (note that adding a hyphen after put instructs the put command to stream data from standard input to hdfs): shell$ hadoop fs -cat people.txt | awk -f"," '{ print $1","$2","$3$4$5 }' | hadoop fs -put - people-coalesed.txt btw, if your input and output files are lzop-compressed then this command would work: shell$ hadoop fs -cat people.txt.lzo | lzop -dc | awk -f"," '{ print $1","$2","$3$4$5 }' | \ lzop -c | hadoop fs -put - people-coalesed.txt.lzo this is great if your file isn’t too large, but if it’s multiple gigabytes in length then you probably want to harness the power of mapreduce to get this done in a jiffy! the words “in a jiffy” and “mapreduce” aren’t commonly used together, so what do we do? well you could crack open pig or hive and write some custom user-defined functions, but this means you end up in java which we want to avoid. hadoop streaming comes to the rescue in these situations. let’s first create our awk script which will be executed: shell$ cat people.awk #!/bin/awk -f begin { fs = "," } { print $1","$2","$3$4$5 } in linux, if you make this awk script executable, you could execute is as follows: shell$ ./people.awk people.txt in mapreduce-land we don’t need to join data in this particular example, so we don’t need to run any reducers. call your awk script from mappers via hadoop streaming with this command: shell$ hadoop_home=/usr/lib/hadoop shell$ ${hadoop_home}/bin/hadoop \ jar ${hadoop_home}/contrib/streaming/*.jar \ -d mapreduce.job.reduces=0 \ -d mapred.reduce.tasks=0 \ -input people.txt \ -output people-coalesed \ -mapper people.awk \ -file people.awk a few options in the hadoop streaming command are worth examining: finally - to get lzo into the picture you need to add -inputformat , -d mapred.output.compress and -d mapred.output.compression.codec arguments: shell$ hadoop_home=/usr/lib/hadoop shell$ ${hadoop_home}/bin/hadoop \ jar ${hadoop_home}/contrib/streaming/*.jar \ -d mapreduce.job.reduces=0 \ -d mapred.reduce.tasks=0 \ -d mapred.output.compress=true \ -d stream.map.input.ignorekey=true \ -d mapred.output.compression.codec=com.hadoop.compression.lzo.lzopcodec \ -inputformat com.hadoop.mapred.deprecatedlzotextinputformat \ -input people.txt.lzo \ -output people-coalesed \ -mapper people.awk \ -file people.awk