Befriending Dragons

Turn Scary Into Attainable


3 Comments

Create HDInsight Cluster in Azure Portal

Creating an HDInsight cluster from the Azure portal is very easy. However, sometimes you want all the choices and best practices explained as well as the "how to". I have created a series of slides with audio recordings to walk you through the process and choices. They are available as sessions 1-8 of "Create HDInsight Cluster in Azure Portal" on my YouTube channel Small Bites of Big Data.

Playlist Getting Started with HDInsight: https://www.youtube.com/playlist?list=PLAD2dOpGM3s1R2L5HgPMX4MkTGvSza7gv 

  1. Why HDInsight: https://youtu.be/J9KzIShLeD8
  2. Azure Subscription: https://youtu.be/lSxMtmRE114
  3. Azure Storage – WASB: https://youtu.be/6OdDDmdaVVE
  4. Metastore: https://youtu.be/1Og_eftYVpA
  5. Create HDInsight: https://youtu.be/SysIo3LwONk
  6. Hive Query: https://youtu.be/DRAuOXsuec0
  7. Load Demo Data: https://youtu.be/XyiOpRPjfUs
  8. Pricing, Automation, and Wrapup: https://youtu.be/78YowrOnNGM

PowerPoint deck: http://www.slideshare.net/cindygross1/create-hd-insightfeb2015 

image

Why HDInsight?

HDInsight is Hadoop on Azure as a service.

  • Easy, cost effective, changeable scale out data processing
  • Lower TCO – easily add/remove/scale
  • Separation of storage and compute allows data to exist across clusters
  • Hortonworks HDP is one of the 3 major Hadoop
    distributors, the most purely open source
  • HDInsight *IS* Hortonworks HDP as a service in Azure (cloud)
  • Metastore (Hcatalog) exists independently across clusters via SQL DB
  • #, size, type of clusters are flexible and can all access the same data
  • Hive is a Hadoop component that makes data look like rows/columns for data warehouse type activities

It offers the standard advantages of Hadoop:

  • Scale-out
  • Load data now, add schema later (write once, read many)
  • Fail fast – iterate through many questions to find the right question
  • Faster time from question to insight
  • Hadoop is “just another data source” for BI, Analytics, Machine Learning

In addition you have the advantages of Hadoop in the cloud:

  • Instantly access data born in the cloud
  • Easily, cheaply load, share, and merge public or private data
  • Data exists independently across clusters (separation of storage and compute) via WASB on Azure storage accounts

Recording of why HDInsight on YouTube

Azure Subscription

You have many options to obtain a Microsoft Azure subscription:

 

Login to Azure Subscription

1. Login on Azure Portal https://manage.windowsazure.com

2. Use a Microsoft Account http://www.microsoft.com/en-us/account/default.aspx
Note: Some companies have federated their accounts and can use company accounts.

image

Choose Subscription

Most accounts will only have one Azure subscription associated with them. But if you seem to have unexpected resources, check to make sure you are in the expected subscription. The Subscriptions button is on the upper right of the Azure portal.

image

image

Add Accounts

Option: Add more Microsoft Accounts as admins of the Azure Subscription.

1. Choose SETTINGS at the very bottom on the left.

2. Then choose ADMINISTRATORS at the top. Click on the ADD button at the very bottom.

3. Enter a Microsoft Account or federated enterprise account that will be an admin.

image 

Recording of getting started with an Azure subscription on YouTube

Azure Storage – WASB

I recommend you manually create at least one Azure storage account and container ahead of time. While the HDInsight creation dialogue gives the option of creating the storage account and container for you, that only works if you don’t plan to reuse data across clusters.

Create a Storage Account

1. Click on STORAGE in the left menu then NEW.

2. URL: Choose a lower-case storage account name that is unique within *.core.windows.net.

3. LOCATION: Choose the same location for the SQL Azure metastore database, the storage account(s), and HDInsight.

4. REPLICATION: Locally redundant stores fewer copies and costs less.

image

Repeat if you need additional storage.

Create a Container

1. Click on your storage account in the left menu then CONTAINERS on the top.

2. Choose CREATE A CONTAINER or choose the NEW button at the bottom.

3. Enter a lower-case NAME for the container, unique within that storage account.

4. Choose either Private or Public ACCESS. If there is any chance of sensitive or PII data being loaded to this container choose Private. Private access requires a key. HDInsight can be configured with that key during creation or keys can be passed in for individual jobs.

This will be the default container for the cluster. If you want to manage your data separately you may want to create additional containers.

image

WASB

Additional information about storage, including details on Windows Azure Storage Blobs (WASB) is on http://SmallBitesOfBigData.com.

image

Recording of creating an Azure storage account and container on YouTube.

Metastore (HCatalog)

In Azure you have the option to create a metastore for Hive and/or Oozie that exists independently of your HDInsight clusters. This allows you to reuse your Hive schemas and Oozie workflows as you drop and recreate your cluster(s). I highly recommend using this option for a production environment or anything that involves repeated access to the same, standard schemas and/or workflows.

Create a Metastore aka Azure SQL DB

Persist your Hive and Oozie metadata across cluster instances, even if no cluster exists, with an HCatalog metastore in an Azure SQL Database. This database should not be used for anything else. While it works to share a single metastore across multiple instances it is not officially tested or supported.

1. Click on SQL DATABASES then NEW and choose CUSTOM CREATE.

2. Choose a NAME unique to your server.

3. Click on the “?” to help you decide what TIER of database to create.

4. Use the default database COLLATION.

5. If you choose an existing SERVER you will share sysadmin access with other databases.

image

You can make the system more secure if you create a custom login on the Azure server. Add that login as a user in the database you just created. Grant it minimal read/write permissions in the database. This is not well documented or tested so the exact permissions needed for this are vague. You may see odd errors if you don’t grant the appropriate permissions.

Firewall Rules

In order to refer to the metastore from automated cluster creation scripts such as PowerShell your workstation must be added to the firewall rules.

1. Click on MANAGE then choose YES.

2. You can also use the MANAGE button to connect to the SQL Azure database and manage logins and permissions.

image

Recording of creating the metastore on YouTube.

Create the HDInsight Cluster

Now that we have the pre-requisites done we can move on to creating the cluster.

  • Quick Create through the Azure portal is the fastest way to get started with all the default settings.
  • The Azure portal Custom Create allows you to customize size, storage, and other configuration options.
  • You can customize and automate through code including .NET and PowerShell. This increases standardization and lets you automate the creation and deletion of clusters over time.
  • For all the examples here we will create a basic Hadoop cluster with Hive, Pig, and MapReduce.
  • A cluster will take several minutes to create, the type and size of the cluster have little impact on the time for creation.

Quick Create Option

For your first cluster choose a Quick Create.

1. Click on HDINSIGHT in the left menu, then NEW.

2. Choose Hadoop. HBase and Storm also include the features of a basic Hadoop cluster but are optimized for in-memory key value pairs (HBase) or alerting (Storm).

3. Choose a NAME unique in the azurehdinisght.net domain.

4. Start with a small CLUSTER SIZE, often 2 or 4 nodes.

5. Choose the admin PASSWORD.

6. The location of the STORAGE ACCOUNT determines the location of the cluster.

image

Custom Create Option

You can also customize your size, admin account, storage, metastore, and more through the portal. We’ll walk through a basic Hadoop cluster.

New

1. Click on HDINSIGHT in the left menu, then NEW in the lower left.

2. Choose CUSTOM CREATE.

image

Basic Info

1. Choose a NAME unique in the azurehdinisght.net domain.

2. Choose Hadoop. HBase and Storm also include the features of a basic Hadoop cluster but are optimized for in-memory key-value pairs (HBase) or alerting (Storm).

3. Choose Windows or Linux as the OPERATING SYSTEM. Linux is only available if you have signed up for the preview.

4. In most cases you will want the default VERSION.

image

Size and Location

1. Choose the number of DATA NODES for this cluster. Head nodes and gateway nodes will also be created and they all use HDInsight cores. For information on how many cores are used by each node see the “Pricing details” link.

2. Each subscription has a billing limit set for the maximum number of HDInsight cores available to that subscription. To change the number available to your subscription choose “Create a support ticket.” If the total of all HDInsight cores in use plus the number needed for the cluster you are creating exceeds the billing limit you will receive a message: “This cluster requires X cores, but only Y cores are available for this subscription”. Note that the messages are in cores and your configuration is specified in nodes.

3. The storage account(s), metastore, and cluster will all be in the same REGION.

image

Cluster Admin

1. Choose an administrator USER NAME. It is more secure to avoid “admin” and to choose a relatively obscure name. This account will be added to the cluster and doesn’t have to match any existing external accounts.

2. Choose a strong PASSWORD of at least 10 characters with upper/lower case letters, a number, and a special character. Some special characters may not be accepted.

image

Metastore (HCatalog)

On the same page as the Hadoop cluster admin account you can optionally choose to use a common metastore (Hcatalog).

1. Click on the blue box to the right of “Enter the Hive/Oozie Metastore”. This makes more fields available.

2. Choose the SQL Azure database you created earlier as the METASTORE.

3. Enter a login (DATABASE USER) and PASSWORD that allow you to access the METASTORE database. If you encounter errors, try logging in to the database manually from the portal. You may need to open firewall ports or change permissions.

image

Default Storage Account

Every cluster has a default storage account. You can optionally specify additional storage accounts at cluster create time or at run time.

1. To access existing data on an existing STORAGE ACCOUNT, choose “Use Existing Storage”.

2. Specify the NAME of the existing storage account.

3. Choose a DEFAULT CONTAINER on the default storage account. Other containers (units of data management) can be used as long as the storage account is known to the cluster.

4. To add ADDITIONAL STORAGE ACCOUNTS that will be accessible without the user providing the storage account key, specify that here.

image

Additional Storage Accounts

If you specified there will be additional accounts you will see this screen.

1. If you choose “Use Existing Storage” you simply enter the NAME of the storage account.

2. If you choose “Use Storage From Another Subscription” you specify the NAME and the GUID KEY for that storage account.

image image

Script Actions

You can add additional components or configure existing components as the cluster is deployed. This is beyond the scope of this demo.

1. Click “add script action” to show the remaining parameters.

2. Enter a unique NAME for your action.

3. The SCRIPT URI points to code for your custom action.

4. Choose the NODE TYPE for deployment.

image

Create is Done!

Once you click on the final checkmark Azure goes to work and creates the cluster. This takes several minutes. When the cluster is ready you can view it in the portal.

image

Recording of HDInsight quick and custom create on YouTube

Query with Hive

For most people the easiest, fastest way to learn Hadoop is through Hive. Hive is also the most widely used component of Hadoop. When you use the Hive ODBC driver any ODBC-compliant app can access the Hive data as "just another data source". That includes Azure Machine Learning, Power BI, Excel, and Tableau.

Hive Console

The simplest, most relatable way for most people to use Hadoop is via the SQL-like, Database-like Hive and HiveQL (HQL).

1.  Put focus on your HDInsight cluster and choose QUERY CONSOLE to open a new tab in your browser. In my case it opens: https://dragondemo1.azurehdinsight.net//

2.  Click on Hive Editor.

image

image

Query Hive

The query console defaults to selecting the first 10 rows from the pre-loaded sample table. This table is created when the cluster is created.

1. Optionally edit or replace the default query:
Select * from hivesampletable LIMIT 10;

2. Optionally name your query to make it easier to find in the job history.

3. Click Submit.

Hive is a batch system optimized for processing huge amounts of data. It spends several seconds up front splitting the job across the nodes and this overhead exists even for small result sets. If you are doing the equivalent of a table scan in SQL Server and have enough nodes in Hadoop, Hadoop will probably be faster than SQL Server. If your query uses indexes in SQL Server, then SQL Server will likely be faster than Hive.

image

View Hive Results

1. Click on the Query you just submitted in the Job Session. This opens a new tab.

image

2. You can see the text of the Job Query that was submitted. You can Download it.

3. The first few lines of the Job Output (query result) are available. To see the full output choose Download File.

4. The Job Log has details including errors if there are any.

5. Additional information about the job is available in the upper right.

image

View Hive Data in Excel Workbook

At this point HDInsight is “just another data source” for any application that supports ODBC.

1. Install the Microsoft Hive ODBC driver.

2. Define an ODBC data source pointing to your HDInsight instance.

3. From DATA choose From Other Sources and From Data Connection Wizard.

image

View Hive Data in PowerPivot

At this point HDInsight is “just another data source” for any application that supports ODBC.

1. Install the Microsoft Hive ODBC driver.

2. Define an ODBC data source pointing to your HDInsight instance.

3. Click on POWERPIVOT then choose Manage. This opens a new PowerPivot for Excel window.

4. Choose Get External Data then Others (OLEDB/ODBC).

Now you can combine the Hive data with other data inside the tabular PowerPivot data model.

image

Recording of querying Hive on YouTube

Load Demo Data

In the cloud you don’t have to load data to Hadoop, you can load data to an Azure Storage Account. Then you point your HDInsight or other WASB compliant Hadoop cluster to the existing data source. There many ways to load data, for the demo we’ll use CloudXplorer.

You use the Accounts button to add Azure, S3, or other data/storage accounts you want to manage.

In this example nealhadoop is the Azure storage account, demo is the container, and bacon is a “directory”. The files are bacon1.txt and bacon2.txt. Any Hive tables would point to the bacon directory, not to individual files. Drag and drop files from Windows Explorer to CloudXplorer.

Windows Azure Storage Explorers (2014)

image

Recording of loading demo data on YouTube

WrapUp

Once you have created the HDInsight cluster you can use it and play with it and try many things. When you are done, simply remove the cluster. If you created an independent metastore in SQL Azure you can use that same metastore and the same Azure storage account(s) the next time you create a cluster. You are charged for the existence of the cluster, not for the usage of it. So make sure you drop the cluster when you aren’t using it. You can use automation, such as PowerShell, to spin up a cluster that is configured the same every time and to drop it. Check the website for the most recent information.

Pricing

image

Automate with PowerShell

With PowerShell, .NET, or the Cross-Platform cmd line tools you can specify even more configuration settings that aren’t available in the portal. This includes node size, a library store, and changing default configuration settings such as Tez and compression.

Automation allows you to standardize and with version control lets you track your configurations over time.

Sample PowerShell Script: HDInsight Custom Create http://blogs.msdn.com/b/cindygross/archive/2013/12/06/sample-powershell-script-hdinsight-custom-create.aspx. If your HDInsight and/or Azure cmdlets don’t match the current documention or return unexpected errors run Web Platform Installer and check for a new version of “Microsoft Azure PowerShell with Microsoft Azure SDK” or “Microsoft Azure PowerShell (standalone).”

image

Recording of Pricing, Automation, and Wrapup on YouTube

Summary

  • HDInsight is Hadoop on Azure as a service, specifically Hortonworks HDP on either Windows or Linux
  • Easy, cost effective, changeable scale out data processing for a lower TCO – easily add/remove/scale
  • Separation of storage and compute allows data to exist across clusters via WASB
  • Metastore (Hcatalog) exists independently across clusters via SQL DB
  • #, size, type of clusters are flexible and can all access the same data
  • Instantly access data born in the cloud; Easily, cheaply load, share, and merge public or private data
  • Load data now, add schema later (write once, read many)
  • Fail fast – iterate through many questions to find the right question
  • Faster time from question to insight
  • Hadoop is “just another data source” for BI, Analytics, Machine Learning

I hope you enjoyed this Small Bite of Big Data! Happy Hadooping!

Cindy Gross – Neal Analytics: Big Data and Cloud Technical Fellow  
@SQLCindy | @NealAnalytics | CindyG@NealAnalytics.com | http://smallbitesofbigdata.com

Advertisements


3 Comments

Taking Flight a.k.a. The Data Dragon’s Life After Microsoft

Cross-posted (with slightly worse formatting) from http://befriendingdragons.com/2014/07/23/taking-flight-a-k-a-the-data-dragons-life-after-microsoft/

Taking flight like Toothless from How to Train Your DragonLife is a journey – we can choose to fly through it with our wings spread to catch and channel the winds, or we can let the winds pummel us to the ground. I choose to take flight, enjoy the journey, and land on my feet. Then take off again. Even when the flight happens because of an unexpected push from the nice, comfy nest, it’s possible to spread our wings and take off in the direction we choose. Especially when you’ve decided you’re a Data Dragon. Yes, that’s me. Cindy the Data Dragon.

Wha...? Huh? Wha…? Huh?

What am I talking about? One of those life changing events that sneaks up on you sometimes.

Last Thursday I got a very unexpected call and I got to experience hearing the words “you’ve been laid off” for the first time ever. It was effective the same day, at least as far as job elimination. I am a Microsoft employee until September 15, my options are wide open after that.

I could choose to sit around and feel sorry for myself, ask countless “why me” and “why now” questions. What I did instead is remember that I am likely in a far better position than many of the other 13,000 people laid off the same day. And remember that now I don’t have to wonder and worry about the remaining Microsoft layoffs that are expected. And remember that this opens up many wonderful opportunities. And remember all the friends, co-workers, and customers who instantly offered support (thanks Sean, Terry, and Linda for the coffee followed by the much stronger drink and the rest of you for all the calls, emails, and IMs). And thank those same folks for the job leads, introductions, and recommendations on LinkedIn that immediately started pouring in – keep them coming! The Data Dragon chooses to concentrate on the good things, dive into making sense of things, and move on to new and better things. (Yes, Murshed, I again referred to myself in the 3rd person).

So now what?

I am going camping this week, I plan to make time for getting out of town again for a few days or weeks before the end of September (SLC ComicCon anyone?), and I am going to get my beautiful back yard back under control and add more colorful things growing in it. I am going to take my time finding the right Big Data job, not just any job.

Connect with me on Skype (cindygross@outlook.com), follow me on Twitter (SQLCindy | Cindygross), and send pics of you toasting the Data Dragon and her beautiful future!

Don’t stand in my way, the Data Dragon is taking flight and looking forward to all the wonderful things in my future!

Green-eyed Data Dragons like me never stay down long!

Green-eyed Data Dragons like me never stay down long!


16 Comments

Taking Flight a.k.a. The Data Dragon’s Life After Microsoft

Taking flight like Toothless from How to Train Your Dragon

Taking flight like Toothless from How to Train Your Dragon

Life is a journey – we can choose to fly through it with our wings spread to catch and channel the winds, or we can let the winds pummel us to the ground. I choose to take flight, enjoy the journey, and land on my feet. Then take off again. Even when the flight happens because of an unexpected push from the nice, comfy nest, it’s possible to spread our wings and take off in the direction we choose. Especially when you’ve decided you’re a Data Dragon. Yes, that’s me. Cindy the Data Dragon.

Wha...? Huh?

Wha…? Huh?

What am I talking about? One of those life changing events that sneaks up on you sometimes.

Last Thursday I got a very unexpected call and I got to experience hearing the words “you’ve been laid off” for the first time ever. It was effective the same day, at least as far as job elimination. I am a Microsoft employee until September 15, my options are wide open after that.

I could choose to sit around and feel sorry for myself, ask countless “why me” and “why now” questions. What I did instead is remember that I am likely in a far better position than many of the other 13,000 people laid off the same day. And remember that now I don’t have to wonder and worry about the remaining Microsoft layoffs that are expected. And remember that this opens up many wonderful opportunities. And remember all the friends, co-workers, and customers who instantly offered support (thanks Sean, Terry, and Linda for the coffee followed by the much stronger drink and the rest of you for all the calls, emails, and IMs). And thank those same folks for the job leads, introductions, and recommendations on LinkedIn that immediately started pouring in – keep them coming! The Data Dragon chooses to concentrate on the good things, dive into making sense of things, and move on to new and better things. (Yes, Murshed, I again referred to myself in the 3rd person).

So now what?

I am going camping this week, I plan to make time for getting out of town again for a few days or weeks before the end of September (SLC ComicCon anyone?), and I am going to get my beautiful back yard back under control and add more colorful things growing in it. I am going to take my time finding the right Big Data job, not just any job.

Connect with me on Skype (cindygross@outlook.com), follow me on Twitter (SQLCindy | Cindygross), and send pics of you toasting the Data Dragon and her beautiful future!

Don’t stand in my way, the Data Dragon is taking flight and looking forward to all the wonderful things in my future!

Green-eyed Data Dragons like me never stay down long!

Green-eyed Data Dragons like me never stay down long!


Leave a comment

Use Additional Storage Accounts with HDInsight Hive

When you create an HDInsight Hadoop cluster you pass in one or more storage accounts and their associated keys. This allows you to access the files on all associated storage accounts from the cluster. If you want to use public storage that isn’t passed in at create time that’s easy – simply supply the storage account name each time you run a job. But how do you access data on private storage accounts that need an access key?

The steps are laid out in this wiki by Eric Hanson: Using an HDInsight Cluster with Alternate Storage Accounts and Metastores

http://social.technet.microsoft.com/wiki/contents/articles/23256.using-an-hdinsight-cluster-with-alternate-storage-accounts-and-metastores.aspx 

I am providing a variable based variation of the PowerShell sample for Hive. To set up PowerShell for use with Azure see Getting Started with Azure PowerShell Cmdlets–Subscription Management.

First you will set some values for your environment. If you use your default subscription you don’t need to pass in the subscription name and select it. However, you will always need to specify the HDInsight cluster name. In this example $undefinedStorageAccount is the name of an account that you want to access from a cluster but you didn’t define it when you created the cluster. You always need to specify which container to use for any given reference so you also need to define $undefinedContainer. If the storage account belongs to the current subscription you can simply ask Azure to return the key (#commented out in the example below) or you can paste in the key that someone has given you.

$subscriptionName = "LocalAzureSubscriptionName"
$clusterName = "HDInsightClusterName"
$undefinedStorageAccount = "AdditionalStorageAccount"
$undefinedContainer = "ContainerOnAdditionalStorageAccount"
#$undefinedStorageKey = Get-AzureStorageKey $undefinedStorageAccount | %{ $_.Primary }
$undefinedStorageKey = "YourActualAccessKeyFromAzurePortal"

Now choose which of your locally defined subscriptions to use:

Select-AzureSubscription -SubscriptionName $subscriptionName

Set the context of the cluster you want to use:

Use-AzureHDInsightCluster $clusterName

Now let’s check your HDInsight cluster properties.

$defaultStorageAccount  = (Get-AzureHDInsightCluster -Name $clusterName).DefaultStorageAccount.StorageAccountName #default/only storage account
$defaultContainerName   = (Get-AzureHDInsightCluster -Subscription $SubID -Cluster $ClusterName).DefaultStorageAccount.StorageContainerName
$definedStorageAccounts = (Get-AzureHDInsightCluster -Name $clusterName).StorageAccounts #no 2nd account is associated, no value is returned

Let’s check the values and verify that the storage account you want to use is not listed as either the DefaultStorageAccount (every cluster has one) or as one of the additional known storage accounts configured during provisioning (you may have zero, one, or many).

write-host "===Default storage account"
$defaultStorageAccount
write-host "===Default container name"
$defaultContainerName
write-host "===Other defined storage accounts for this cluster"
$definedStorageAccounts

Next we’ll get a non-recursive listing of the files in the default location:

invoke-hive "dfs -ls wasb://$defaultContainerName@$defaultStorageAccount/;" #default storage

And then try to get a listing for the private storage account that we have not associated with the cluster:

invoke-hive "dfs -ls wasb://$undefinedContainer@$undefinedStorageAccount/;" #not associated, errors

Because the storage account access key is not yet known you will see an error similar to this one:

Logging initialized using configuration in file:/C:/apps/dist/hive-0.12.0.2.0.7.0-1559/conf/hive-log4j.properties
ls: org.apache.hadoop.fs.azure.AzureException: Unable to access container xyz in account abc using anonymous credentials, 
and no credentials found for them  in the configuration.
Command failed with exit code = 1

But we can fix this! From PowerShell we can pass in “defines” statements to change configuration values, add libraries, etc.

$defines = @{}
$defines.Add("fs.azure.account.key.$undefinedStorageAccount.blob.core.windows.net", $undefinedStorageKey)
Invoke-Hive -Defines $defines -Query "dfs -ls wasb://$undefinedContainer@$undefinedStorageAccount.blob.core.windows.net/;"

The access key is only available to this Hive query, but now that I have the variables set I can pass it in to other queries as well. Happy Hiving!

I hope you enjoyed this small bite of Big Data!


1 Comment

HDInsight: Jiving about Hadoop and Hive with CAT

Tomorrow I will be talking about Hive as part of Pragmatic Work’s Women in Technology (WIT) month of webcasts. I am proud to be part of this lineup with all these stellar WITs! I encourage my fellow WITs to get more involved in your data community and if you don’t already do so start tweeting, blogging, and speaking. I am happy to coach you through your first speaking engagement if you are interested. Get out there and start showing the world what you can do!

Thursday’s talk is going to be HDInsight: Jiving about Hadoop and Hive with CAT. Let’s break that title down.

HDInsight is Microsoft’s distribution of Hadoop. As part of the HDInsight project we have checked code back into the core Apache Hadoop source code to make the core code runs great on Windows. We are also adding functionality and features such as JavaScript and Azure Storage Vault that make the product more robust and enterprise friendly. This week the HDInsight Service Preview on Azure became available to those with an Azure subscription.

Hadoop is a scale out methodology that allows businesses to quickly consume and analyze data in ways they haven’t been able to before. This can lead to faster, better business insights and business actions.

Hive is a way to impose metadata and structure on the loosely structured (unstructured, multi-structured, semi-structured) data that resides in Hadoop’s HDFS file system. With Hive and the Hive ODBC driver you can make Hadoop data look like any other data source to your familiar BI tools such as Excel. PowerPivot can connect to Hive data, mash that data up with existing data sources such as SQL Azure, SQL Server, and OData, and allow you to visualize it with Power View. I have an end to end demo of this: Hurricane Sandy Mash-Up: Hive, SQL Server, PowerPivot & Power View.

CAT is my team at Microsoft. The Customer Advisory Team (CAT) works with customers who are doing new, unusual, and interesting things that push the boundaries of technology. We share what we learn with the community so you can do your jobs better and we take what we learn from you to the product team to help improve the product.

My slides are attached at the bottom of this post. I believe a recording of the talk will be posted by Pragmatic Works on their site.

I look forward to “seeing” you all at my talk tomorrow and would love to see your tweets or hear directly from you afterwards.

I hope you’ve enjoyed this small bite of big data! Look for more blog posts soon on the samples and other activities.

Note: the CTP and TAP programs are available for a limited time. Details of the usage and the availability of the CTP may change rapidly.

Digg This

PragmaticWorksHDInsightJivingAboutHadoopAndHiveWithCATMar202013.pptx


11 Comments

HDInsight: Hive Internal and External Tables Intro

Small Bites of Big Data

Cindy Gross, SQLCAT PM

HDInsight is Microsoft’s distribution, in partnership with Hortonworks, of Hadoop. Hive is the component of the Hadoop ecosystem that imposes structure on Hadoop data in a way that makes it usable from BI tools that expect rows and columns with defined data types. Hive tables can be created as EXTERNAL or INTERNAL. This is a choice that affects how data is loaded, controlled, and managed.

Use EXTERNAL tables when:

  • The data is also used outside of Hive. For example, the data files are read and processed by an existing program that doesn’t lock the files.
  • Data needs to remain in the underlying location even after a DROP TABLE. This can apply if you are pointing multiple schemas (tables or views) at a single data set or if you are iterating through various possible schemas.
  • You want to use a custom location such as ASV.
  • Hive should not own data and control settings, dirs, etc., you have another program or process that will do those things.
  • You are not creating table based on existing table (AS SELECT).

Use INTERNAL tables when:

  • The data is temporary.
  • You want Hive to completely manage the lifecycle of the table and data.

We’ll walk through creating basic tables with a few rows of data so you can see some of the differences between EXTERNAL and INTERNAL tables. The demo data files are attached at the bottom of the blog. Alternatively you can simply open notepad and create your own files with a series of single column rows. If you create your own files make sure you have a carriage return/line feed at the end of all rows including the last one. The files should be in a Windows directory called c:data on the HDInsight Head Node. For HDInsight Server (on-premises) that’s the machine where you ran setup. For HDInsight Services (Azure) you can create a Remote Desktop connection (RDP) to the head node from the Hadoop portal.

Note: Your client tool editor or the website may change the dashes or other characters in the following commands to “smart” characters. If you get syntax errors from a direct cut/paste, try pasting into notepad first or deleting then retyping the dash (or other special characters).

Create an HDInsight cluster. You can do this on your own Windows machine by installing HDInsight Server or by signing up for HDInsight Services on Azure. For the CTP of HDInsight Services as of February 2013 you fill out a form to request access and receive access within a few days. Soon the service will be available from the Azure portal via your Azure subscription. Since the portal interface will be changing soon and all the commands are straightforward I will show you how to do all the steps through the Hive CLI (command line interface).

Open a Hadoop Command Prompt:

HadoopCmdPrompt

Change to the Hive directory (necessary in early preview builds of Hive):

cd %hive_home%bin

Load some data (hadoop file system put) and then verify it loaded (hadoop file system list recursively):

hadoop fs -put c:databacon.txt /user/demo/food/bacon.txt

hadoop fs -lsr /user/demo/food

The put command doesn’t return a result, the list command returns one row per file or subdirectory/file:

-rw-r–r–   1 cgross supergroup        124 2013-02-05 22:41 /user/demo/food/bacon.txt

Enter the Hive CLI (command line interface):

hive

Tell Hive to show the column names above the results (all Hive commands require a semi-colon as a terminator, no result is returned from this set command):

Set hive.cli.print.header=true;

Create an INTERNAL table in Hive and point it to the directory with the bacon.txt file:

CREATE INTERNAL TABLE internal1 (col1 string) LOCATION ‘/user/demo/food’;

Oops… that failed because INTERNAL isn’t a keyword, the absence of EXTERNAL makes it a managed, or internal, table.

FAILED: Parse Error: line 1:7 Failed to recognize predicate ‘INTERNAL’.

So let’s create it without the invalid INTERNAL keyword. Normally we would let an INTERNAL table default to the default location of /hive/warehouse but it is possible to specify a particular directory:

CREATE TABLE internal1 (col1 string) LOCATION ‘/user/demo/food’;

That will return the time taken but no other result. Now let’s look at the schema that was created:. Note that the table type is MANAGED_TABLE.

DESCRIBE FORMATTED internal1;

col_name        data_type       comment
# col_name              data_type               comment

 

col1                    string                  None

 

# Detailed Table Information
Database:               default
Owner:                  cgross
CreateTime:             Tue Feb 05 22:45:57 PST 2013
LastAccessTime:         UNKNOWN
Protect Mode:           None
Retention:              0
Location:               hdfs://localhost:8020/user/demo/food
Table Type:             MANAGED_TABLE
Table Parameters:
        transient_lastDdlTime   1360133157

 

# Storage Information
SerDe Library:          org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe
InputFormat:            org.apache.hadoop.mapred.TextInputFormat
OutputFormat:           org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat
Compressed:             No
Num Buckets:            -1
Bucket Columns:         []
Sort Columns:           []
Storage Desc Params:
        serialization.format    1

And now look at some rows:

SELECT * FROM internal1;

col1
HDInsight_Bacon
SQL_Bacon
PASS_bacon
Summit_BACON
Database_Bacon
NoSQL_Bacon
BigData_Bacon
Hadoop_Bacon
Hive_Bacon

What happens if we don’t specify a directory for an INTERNAL table?

CREATE TABLE internaldefault (col1 string);

It is created in the default Hive directory, which by default is in /hive/warehouse (dfs shells back out to Hadoop fs):

dfs -lsr /hive/warehouse;

We can see that Hive has created a subdirectory with the same name as the table. If we were to load data into the table Hive would put it in this directory:
drwxr-xr-x   – cgross supergroup          0 2013-02-05 22:52 /hive/warehouse/internaldefault

However, we won’t use this table for the rest of the demo so let’s drop it to avoid confusion. The drop also removes the subdirectory.

DROP TABLE internaldefault;

dfs -lsr /hive/warehouse;

Once we dropped the internaldefault table the directory that Hive created was automatically cleaned up. Now let’s add a 2nd file to the first internal table and check that it exists:

dfs -put c:databacon2.txt /user/demo/food/bacon2.txt;

dfs -lsr /user/demo/food;

-rw-r–r–   1 cgross supergroup        124 2013-02-05 23:04 /user/demo/food/bacon.txt
-rw-r–r–   1 cgross supergroup         31 2013-02-05 23:03 /user/demo/food/bacon2.txt

Since the CREATE TABLE statement points to a directory rather than a single file any new files added to the directory are immediately visible (remember that the column name col1 is only showing up because we enabled showing headers in the output – there is no row value of col1 in the data as headers are not generally included in Hadoop data):

SELECT * FROM internal1;

col1
HDInsight_Bacon
SQL_Bacon
PASS_bacon
Summit_BACON
Database_Bacon
NoSQL_Bacon
BigData_Bacon
Hadoop_Bacon
Hive_Bacon
More_BaCoN
AndEvenMore_bAcOn

Now let’s create an EXTERNAL table that points to the same directory and look at the schema:

CREATE EXTERNAL TABLE external1 (colE1 string) LOCATION ‘/user/demo/food’;

DESCRIBE FORMATTED external1;

col_name        data_type       comment
# col_name              data_type               comment

 

cole1                   string                  None

 

# Detailed Table Information
Database:               default
Owner:                  cgross
CreateTime:             Tue Feb 05 23:07:12 PST 2013
LastAccessTime:         UNKNOWN
Protect Mode:           None
Retention:              0
Location:               hdfs://localhost:8020/user/demo/food
Table Type:             EXTERNAL_TABLE

Table Parameters:
        EXTERNAL                TRUE
        transient_lastDdlTime   1360134432

 

# Storage Information
SerDe Library:          org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe
InputFormat:            org.apache.hadoop.mapred.TextInputFormat
OutputFormat:           org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat
Compressed:             No
Num Buckets:            -1
Bucket Columns:         []
Sort Columns:           []
Storage Desc Params:
        serialization.format    1

This time the table type is EXTERNAL_TABLE. You can see that the location was expanded to include the default settings which in this case are the localhost machine using the default HDFS (as opposed to ASV or Azure Storage Vault).

Now look at the data:

SELECT * FROM external1;

The result set is a combination of the two bacon files:

HDInsight_Bacon
SQL_Bacon
PASS_bacon
Summit_BACON
Database_Bacon
NoSQL_Bacon
BigData_Bacon
Hadoop_Bacon
Hive_Bacon
More_BaCoN
AndEvenMore_bAcOn

That table returns the same data as the first table – we have two tables pointing at the same data set! We can add another one if we want:

CREATE EXTERNAL TABLE external2 (colE2 string) LOCATION ‘/user/demo/food’;

DESCRIBE FORMATTED external2;

SELECT * FROM external2;

You may create multiple tables for the same data set if you are experimenting with various structures/schemas.

Add another data file to the same directory and see how it’s visible to all the tables that point to that directory:

dfs -put c:dataveggies.txt /user/demo/food/veggies.txt;

SELECT * FROM internal1;

SELECT * FROM external1;

SELECT * FROM external2;

Each table will return the same results:

HDInsight_Bacon
SQL_Bacon
PASS_bacon
Summit_BACON
Database_Bacon
NoSQL_Bacon
BigData_Bacon
Hadoop_Bacon
Hive_Bacon
More_BaCoN
AndEvenMore_bAcOn
SQL_Apple
NoSQL_Pear
SQLFamily_Kiwi
Summit_Mango
HDInsight_Watermelon
SQLSat_Strawberries
Raspberrylimelemonorangecherryblueberry 123 456

Now drop the INTERNAL table and then look at the data from the EXTERNAL tables which now return only the column name:

DROP TABLE internal1;

SELECT * FROM external1;

SELECT * FROM external2;

dfs -lsr /user/demo/food;

Result: lsr: Cannot access /user/demo/food: No such file or directory.

Because the INTERNAL (managed) table is under Hive’s control, when the INTERNAL table was dropped it removed the underlying data. The other tables that point to that same data now return no rows even though they still exist!

Clean up the demo tables and directory:

DROP TABLE external1;

DROP TABLE external2;

exit;

This should give you a very introductory level understanding of some of the key differences between INTERNAL and EXTERNAL Hive tables. If you want full control of the data loading and management process, use the EXTERNAL keyword when you create the table.

I hope you’ve enjoyed this small bite of big data! Look for more blog posts soon on the samples and other activities.

Note: the CTP and TAP programs are available for a limited time. Details of the usage and the availability of the TAP and CTP builds may change rapidly.

bacon.zip


4 Comments

Hurricane Sandy Mash-Up: Hive, SQL Server, PowerPivot & Power View

Small Bites of Big Data

                    

Authors: Cindy Gross Microsoft SQLCAT PM, Ed Katibah Microsoft SQLCAT PM

Tech Reviewers: Bob Beauchemin Developer Skills Partner at SQLSkills, Jeannine Nelson-Takaki Microsoft Technical Writer, John Sirmon Microsoft SQLCAT PM, Lara Rubbelke Microsoft Technical Architect, Murshed Zaman Microsoft SQLCAT PM

For my #SQLPASS Summit 2012 talk SQLCAT: Big Data – All Abuzz About Hive (slides available to all | recording available to PASS Summit 2012 attendees) I showed a mash-up of Hive, SQL Server, and Excel data that had been imported to PowerPivot and then displayed via Power View in Excel 2013 (using the new SharePoint-free self-service option). PowerPivot brings together the new world of unstructured data from Hadoop with structured data from more traditional relational and multi-dimensional sources to gain new business insights and break down data silos. We were able to take very recent data from Hurricane Sandy, which occurred the week before the PASS Summit, and quickly build a report to pinpoint some initial areas of interest. The report provides a sample foundation for exploring to find additional insights. If you need more background on Big Data, Hadoop, and Hive please see my previous blogs and talks.

I will walk you through the steps to create the report including loading population demographics (census), weather (NOAA), and lookup table (state abbreviations) data into Hive, SQL Server, Excel, and PowerPivot then creating visualizations in Power View to gain additional insights. Our initial goal is to see if there are particular geographic areas in the path of Hurricane Sandy that might need extra assistance with evacuation. One hypothesis is that people above a given age might be more likely to need assistance, so we want to compare age data with the projected rainfall patterns related to the path of the hurricane. Once you see this basic demonstration you can envision all sorts of additional data sets that could add value to the model, along with different questions that could be asked given the existing data sets. Data from the CDC, pet ownership figures, housing details, job statistics, zombie predictions, and public utility data could be added to Hive or pulled directly from existing sources and added to the report to gain additional insights. Those insights might, for example, help first responders during future storms, assist your business to understand various ways it can help after a storm or major cleanup effort, or aid future research into reducing the damage done by natural disasters.

Prerequisites

  • Excel 2013 (a 30 day trial is available)
  • SQL Server 2008 R2 or later
  • HDInsight Server or HDInsight Service (Service access will take a few days so plan in advance or use Server)
  • Hive ODBC Driver (available with HDInsight)
  • NOAA Data Set restored to SQL Server 2008 R2 or later (available from SkyDrive)
  • Census Data Set (attached at bottom of blog)

 

SQL Server

Relational Data

One of the data sets for the demo is NOAA weather data that includes spatial characteristics. Since SQL Server has a rich spatial engine and the data set is well known and highly structured that data was a good fit for SQL Server. Spatial_Ed has a wiki post on how to create the finished data set from raw NOAA data. Ed also provides a backup of the completed data set for our use. Take the SQL Server database backup (download here) made available by Ed in his wiki post and restore it to your SQL Server 2008 R2 or later instance as a database called NOAA.

USE [master];
RESTORE DATABASE [NOAA] FROM  DISK = N'C:DemoBigDataSandyNOAA.bak' 
WITH MOVE N'NOAA2' TO N'C:DATANOAA2.mdf',  
MOVE N'NOAA2_log' TO N'C:DATANOAA2_log.LDF';
GO
 

Since this data will be used by business users, add a view to the NOAA database with a friendlier name for the rainfall/flashflood data:

USE NOAA;
GO
CREATE VIEW flashflood AS SELECT * FROM [dbo].[nws_ffg7];

Take a look at a few rows of the data. For more information on what information is available in the census data, see the U.S. Census Bureau website.

SELECT TOP 10 * FROM flashflood;

Hive

Hive is a part of the Hadoop ecosystem that allows you to create tables and impose structure on Hadoop data. It is available in HDInsight which is Microsoft’s distribution of Hadoop, sometimes referred to as Hadoop on Azure or Hadoop on Windows. HDInsight is currently available for preview in both Azure (HDInsight Service) and on-premises (HDInsight Server) versions. The HDInsight Server version is lightweight and simple to install – you can even put it on your laptop as it is a single node installation for now. Or you can request access to an Azure Hadoop cluster via HDInsight Service, though this takes a few days in the CTP phase. Hive is automatically installed as part of Hadoop in the preview versions, though the Hive ODBC driver requires a separate setup step. Hive can be described in many ways, but for your purposes within this article the key point is that it provides metadata and structure to Hadoop data that allows the data to be treated as “just another data source” to an ODBC compliant application such as Excel. HiveQL, or HQL, looks very similar to other SQL languages and has similar functionality.

ODBC

Once you have access to an HDInsight cluster, install the Hive ODBC driver to your local box (where you have Excel 2013). Make sure you install the ODBC driver platform (32-bit or 64-bit) that matches the platform of Excel. We recommend the 64-bit version of Excel 2013 and of the Hive ODBC driver since PowerPivot is able to take advantage of the larger memory available in x64. The Hive ODBC driver is available from the Downloads tile in the HDInsight Service (Azure) or HDInsight Server (on-premises) portal. Click on the appropriate installer (32-bit or 64-bit) and click through the Microsoft ODBC Driver for Hive Setup. You will end up with version .09 of the Microsoft Hive ODBC driver.

Make sure the Hive ODBC port 10000 is open on your Hadoop cluster (instructions here).

Create a Hive ODBC system DSN pointing to your Hadoop cluster. For this example I used the ODBC Data Source Administrator to create a system DSN called CGrossHOAx64 pointing to my http://HadoopOnAzure.com instance cgross.cloudapp.net with port = 10000 and an account called cgross1. For the on-premise version you can use localhost for the Host value and you will not specify an account or password.

ODBCAdminHive    SystemDSNHive 

Note: With Windows 2008 R2 and earlier or Windows 7 and earlier if you must use 32-bit Excel and the 32-bit Hive ODBC driver (not recommended) on 64-bit Windows, to create the system DSN you have to use the 32-bit version of the ODBC Data Source Administrator. It is in a location similar to C:WindowsSysWOW64odbcad32.exe. In Windows 2012 or later or Windows 8 or later there is a single ODBC Data Source Administrator for both 32-bit and 64-bit drivers.

Key Hadoop Pointers
  • The Hadoop Command Prompt is installed with HDInsight. There will be an icon added to the desktop on your head node. When you open it you will see that it is a basic Windows Command Prompt but with some specific settings applied.
    HadoopCmdPrompt  HadoopCmdPromptOpen
  • In early versions of HDInsight the Hive directory is not in the system path so you have to manually change to the Hive directory (cd %hive_home%bin) to run Hive commands from the Hadoop Command Prompt. This issue will be fixed in future versions.
  • HDFS is the scale-out storage technology that comes as part of core Hadoop. How it works behind the scenes is covered in great detail in many other places and is not relevant to this discussion. This is the default storage in HDInsight Server (on-prem) and the CTP of HDInsight Service (Azure).
  • ASV, or Azure Storage Vault, is the Windows Azure implementation of HDFS that allows HDInsight to use Azure Blob storage. This will be the default for the production HDInsight Service.
Census Data

Census data has a format that can vary quite a bit. The data is collected in different ways in different countries, it may vary over time, and many companies add value to the data and make available extended data sets in various formats. It could contain large amounts of data kept for very long periods of time. In our example the pieces of census data we find useful and how we look at that data (what structures we impose) may change quite a bit as we explore the data. The variety of structures and the flexibility of the ways to look at the data make it a candidate for Hadoop data. Our need to explore the data from common BI tools such as Excel, which expect rows and columns with metadata, leads us to make it into a Hive table.

Census data was chosen because it includes age ranges and this fits with the initial scenario we are building to look at how many older individuals are in various high danger areas in the path of the hurricane. I have attached a demo-sized variation of this data set called Census2010.dat at the bottom of the blog. Download that tab-delimited U.S. census data set to the c:data folder on the head node for your HDInsight cluster. The head node is the machine you installed your single node HDInsight Server (on-prem) on, such as your laptop, or the machine configured for Remote Desktop access from the portal in your HDInsight Service (Azure) cluster. If you wish to explore the raw census data used in this demo, take a look at this site: http://www.census.gov/popest/research/modified.html.

Next, load the census data to Hadoop’s HDFS storage. Unlike a relational system where the structure has to be well-known and pre-defined before the data is loaded, with Hadoop we have the option to load this data into HDFS before we even create the Hive table!

Note: Your client tool editor or the website may change the dashes or other characters in the following commands to “smart” characters. If you get syntax errors from a direct cut/paste, try pasting into notepad first or deleting then retyping the dash (or other special characters).

I will show you three of the many options for loading the data into an external Hive table (in the real world you would have multiple very large files in this directory, but for the purposes of the demo we have one small file), choose any one of the following three options. Note that the directories are created for you automatically as part of the load process. There are many ways to do the load, including cURL, SFTP, PIG, etc. but the steps below are good for illustration purposes.

1) To load the data into HDFS via the Hadoop Command Prompt, open the Hadoop Command Prompt and type:

hadoop fs -put c:datacensus2010.dat /user/demo/census/census.dat

Let’s break down the command into its component pieces. The fs tells Hadoop you have a file system command, put is a data load command, c:datacensus2010.dat is the location of the file within NTFS on Windows, and /user/demo/census/census.dat is the location where we want to put the file within HDFS. Notice that I chose to change the name of the file during the load, mostly just for convenience as the HDFS name I chose is shorter and more generic. By default HDInsight currently defaults to using HDFS so I don’t have to specify that in the command, but if your system has a different default or you just want to be very specific you could specify HDFS in the command by adding hdfs:// before the location.

hadoop fs -put c:datacensus2010.dat hdfs:///user/demo/census/census.dat

Nerd point: You can specify the system name (instead of letting it default to localhost) by appending hdfs://localhost (or a specific name, often remote) to the location. This is rarely done as it is longer and makes the code less portable.

hadoop fs -put c:datacensus2010.dat hdfs://localhost/user/demo/census/census.dat

2) If you are using HDInsight Service (Azure), you can choose to easily store the data in an Azure Blob Store via ASV. If you haven’t already established a connection from your HDInsight cluster to your Azure Blob Store then go into your HDInsight cluster settings and enter your Azure Storage Account name and key. Then instead of letting the data load default to HDFS specify in the last parameter that the data will be loaded to Azure Storage Vault (ASV).

hadoop fs -put c:datacensus2010.dat asv://user/demo/census/census.dat

3) From the JavaScript interactive console in the Hadoop web portal (http://localhost:8085/Cluster/InteractiveJS from the head node) you can use fs.put() and specify either an HDFS or ASV location as your destination (use any ONE of these, not all of them):

/user/demo/census/census.dat
asv://user/demo/census/census.dat
hdfs:///user/demo/census/census.dat

Check from a Hadoop Command Prompt that the file has loaded:

hadoop fs -lsr /user/demo/census

Results will vary a bit, but the output should include the census.dat file: /user/demo/census/census.dat. For example the Hadoop Command Prompt in HDInsight Server should return this row with a different date and time, possibly with a different userid:

-rw-r–r– 1 hadoop supergroup 360058 2013-01-31 17:17 /user/demo/census/census.dat

If you want to run Hadoop file system (fs) commands from the JavaScript console, replace “hadoop fs -” with “#”.

#lsr /user/demo/census

This tells Hadoop that we want to execute a file system command and recursively list the subdirectories and files under the specified directory. If you chose to use ASV then add that prefix (I will assume for the rest of this blog that you know to add asv:/ or hdfs:// if necessary):

hadoop fs -lsr asv://user/demo/census

If you need to remove a file or directory for some reason, you can do so with one of the remove commands (rm = remove a single file or directory, rmr = remove recursively the directory, all subdirectories, and all files therein):

hadoop fs –rm /user/demo/census/census.dat

hadoop fs –rmr /user/demo/census

  

For more details on rm or other commands:

hadoop fs –help

hadoop fs –help rm

hadoop fs –help rmr
  
Hive Table

Create an external Hive table pointing to the Hadoop data set you just loaded. To do this from a command line open your Hadoop Command Prompt on your HDInsight head node. On an Azure cluster, look for the Remote Desktop tile from the HDInsight Service portal. This allows you to connect to the head node and you can optionally choose to save the connection as an RDP file for later use. In the Hadoop Command Prompt, type in the keyword hive to open the interactive Hive command line interface console (CLI). In some early versions of HDInsight the Hive directory isn’t in the system path so you may have to first change to the Hive bin directory in the Hadoop Command Prompt:

cd %hive_home%bin

Now enter the Hive CLI from the Hadoop Command Prompt:

hive

Copy and Paste the below create statement into the Hive CLI window. Or you can paste the CREATE statement into the web portal via the Hive interactive screen. If you are using the ASV option, change the STORED AS line to include asv:/ before the rest of the string (asv://user/demo/census). Take note of the fact that you are specifying the folder (and implicitly all files in it) and not a specific file.

CREATE EXTERNAL TABLE census
(State_FIPS int,
County_FIPS int,
Population bigint,
Pop_Age_Over_69 bigint,
Total_Households bigint,
Median_Household_Income bigint,
KeyID string) 
COMMENT 'US Census Data' 
ROW FORMAT DELIMITED FIELDS TERMINATED BY 't' 
STORED AS TEXTFILE LOCATION '/user/demo/census';

This CREATE command tells Hive to create an EXTERNAL table as opposed to an “internal” or managed table. This means you are in full control of the data loading, data location, and other parameters. Also, the data will NOT be deleted if you drop this table. Next in the command is the column list, each column uses one of the limited number of primitive data types available. The COMMENT is optional, you can also add a COMMENT for one or more of the columns if you like. ROW FORMAT DELIMITED means we have a delimiter between each column and TERMINATED BY ‘t’ means the column delimiter is a tab. The only allowed row delimiter as of Hive 0.9 is a new line and that is the default so we don’t need to specify it. Then we explicitly say this is a collection of TEXTFILEs stored at the LOCATION /user/demo/census (notice that we are not specifying a specific file, but rather the specific directory). We could specify either hdfs:/// or asv:// in front of the location; by default the current default storage location is HDFS so it is an optional prefix (the default can change!).

Now check that the table has been created:

SHOW TABLES;

Check the definition of the table:

DESCRIBE FORMATTED census;

And look at a sampling of the data:

SELECT * FROM census LIMIT 10;

You can now choose to leave the Hive CLI (even the exit requires the semicolon to terminate the command):

EXIT;

Excel 2013 Visualizations

Preparation

In Excel 2013 you don’t have to download PowerPivot or Power View, you just need to make sure they are enabled as COM add-ins. All of the below steps assume you are using Excel 2013.

In Excel 2013 make sure the “Hive for Excel”, “Microsoft Office PowerPivot for Excel 2013”, and “Power View” add-in are enabled. To do this click on the File tab in the ribbon, then choose Options from the menu on the left. This brings up the Excel Options dialog. Choose Add-ins from the menu on the left. At the very bottom of the box you’ll see a section to Manage the Add-ins. Choose “COM Add-ins” and click on the Go… button. This brings up the COM Add-Ins dialog where you can double check that you have checkmarks next to the Hive, PowerPivot, and Power View add-ins. Click OK.

ExcelManageCOMAddins  ExcelCOMAddins

PowerPivot Data

Open a new blank workbook in Excel 2013. We are going to pull some data into PowerPivot, which is an in-memory tabular model engine within Excel. Click on the PowerPivot tab in the ribbon then on the Manage button in the Data Model group on the far left. Click on “Get External Data” in the PowerPivot for Excel dialog.

PowerPivotExternalData

We’re going to mash up data from multiple sources – SQL, Hive, and Excel.

SQL Data

First, let’s get the SQL Server data from the NOAA database. Choose “From Database” then “From SQL Server”. In the Table Import Wizard dialog enter the server and database information:

Friendly connection name = NOAA
Server name = [YourInstanceName such as MyLaptopSQL2012]
Log on to the server = “Use Windows Authentication”
Database name = NOAA

 

ExternalSQLNOAA

Choose “Next >” then “Select from a list of tables and views to choose the data to import”. Put a checkmark next to the flashflood view we created earlier. Click “Finish”. This will import the flash flood data from SQL Server into the in-memory xVelocity storage used by PowerPivot.

ImportFlashFlood

Hive Data

Now let’s load the Hive data. Click on “Get External Data” again. This time choose “From Other Sources” then “Others (OLEDB/ODBC)” and click on the “Next >” button. Enter Census for the friendly name then click on the “Build” button.

Friendly name for this connection = Census
Connection String = “Build”
Provider tab:
   OLE DATABASE Provider(s) = “Microsoft OLE DATABASE Provider for ODBC Drivers”
Connection tab:
   Use data source name = CGrossHOAx64 [the Hive ODBC DSN created above]
   User name = [login for your Hadoop cluster, leave blank for on-premises Nov 2012 preview]
   Password = [password for your Hadoop cluster, leave blank for on-premises Nov 2012 preview]
   Allow saving password = true (checkmark)

 

DSNODBCProvider  DSNHiveAzure

This creates a connection string like this:
Provider=MSDASQL.1;Persist Security Info=True;User ID=cgross1;DSN=CGrossHOAx64;Password=**********

  ExternalHiveCensus2

Choose “Next >” then “Select from a list of tables and views to choose the data to import”. Choose the census table then click on “Finish”. This loads a copy of the Hive data into PowerPivot. Since Hive is a batch mode process which has the overhead of MapReduce job setup and/or HDFS streaming for every query, the in-memory PowerPivot queries against the copy of the data will be faster than queries that go directly against Hive. If you want an updated copy of the data loaded into PowerPivot you can click on the Refresh icon at the top of the PowerPivot window.

Save the Excel file.

Excel Data

Our last data set will be a set of state abbreviations copied into Excel then accessed from PowerPivot. Go back to your Excel workbook (NOT the PowerPivot for Excel window). Rename the first sheet using the tab at the bottom to ShortState (just for clarity, this isn’t used anywhere).

Open https://www.usps.com/send/official-abbreviations.htm in your web browser. Copy the entire “States” table including the headers. Paste the States data into the ShortState worksheet in Excel. Change the headers from State/Possession to State and from Abbreviation to ShortState. Highlight the columns and rows you just added (not the entire column or you will end up with duplicate values). Go to the Insert tab in the ribbon and click on Table in the Tables group on the far left. It will pop up a “Create Table” window, make sure the cells show “=$A$1:$B$60” and that “My table has headers” is checked.

ExternalAbbreviations

Now you will have a Table Tools ribbon at the top on the far right. Click on the Design tab then in the Properties group in the upper left change the Table Name to ShortState. The fact that the worksheet tab, the table, and one column have the same name is irrelevant, that just seemed less confusing to at the time. Now you have an Excel based table called ShortState that is available to be used in the tabular model.

ExternalAbbreviations2

Data Model

Click on the PowerPivot tab and choose Add to Data Model at the top in the Tables group. This will change the focus to the PowerPivot for Excel window. Go to the Home tab in the PowerPivot for Excel window and click on Diagram View in the View group in the upper right. You can resize and rearrange the tables if that makes it easier to visualize the relationships. Add the following relationships by clicking on the relevant column in the first table then dragging the line to the other column in the second table:

Table Flashflood, column State to table ShortState, column State

Table Census, column keyid to table flashflood, column KEYID

PPModelCensus

Click on the flashflood table and then on the Create Hierarchy button in the upper right of the table (as circled in the image above). Name the hierarchy Geography and drag State then County under it.

PPHierarchyCensus

Save the Excel file. You may need to click on “enable content” if you close/reopen the file.

Calculations

In the PowerPivot for Excel window, we’re going to add some calculated columns for later use. Click on the Data View button in the View group at the top. You should now see three tabs at the bottom, one for each of the tables in the model. Click on the census table and then right-click on “Add Column” at the far right of the column headers. Choose to insert column. This will create a new column named CalculatedColumn1. In the formula bar that opened above the column headers paste in “=[Pop_Age_Over_69]/[Population]” (without the quotes). This will populate the new column with values such as 0.0838892880358548. Right-click on the column and choose “Rename Column” – name it PctPopOver69 since it gives us the percentage of the total population that is over 69 years old. That’s the first formula below, repeat those steps for the 2nd and 3rd formulas. PctBuckets bucketizes the output into more meaningful groupings while SortBucketsBy gives us a logical ordering of the buckets. Since most of the numbers fall into a range of less than 20% the buckets are skewed towards that lower range that actually has a variety of data.

  • PctPopOver69=[Pop_Age_Over_69]/[Population]
  • PctBuckets=IF([PctPopOver69]<=.05,”<5%”,IF([PctPopOver69]<.1,”5-10%”, IF([PctPopOver69]<.15,”10-15%”,IF([PctPopOver69]<.2,”15-20%”,”>20%”))))
  • SortBucketsBy=IF([PctPopOver69]<=.05,1,IF([PctPopOver69]<.1,2, IF([PctPopOver69]<.15,3,IF([PctPopOver69]<.2,4,5))))

PPCalculationsCensus

Now let’s use those new calculated columns. Highlight the new PctBuckets column (the 2nd one you just added) and click on the Sort by Column button in the Sort and Filter group. Pick Sort By Column and choose to sort PctBuckets by SortBucketsBy. This affects the order of data in the visualizations we will create later.

PPSortByCensus

Since some of the columns are purely for calculations or IT-type use and would confuse end users as they create visualizations and clutter the field picker, we’re going to hide them. In the Census table right-click on the PctPopOver69 column and choose “Hide from Client Tools”. Repeat for census.SortBucketsBy, census.keyid, census.State_FIPS, census.County_FIPS, flashflood.State_FIPS, flashflood.County_FIPS, and flashflood.KEYID. Our PowerPivot data is now complete so we can close the PowerPivot for Excel window. Note that the Excel workbook only shows the ShortState tab. That’s expected as this is the Excel source data, not the in-memory PowerPivot tables that we’ve created from our three data sources. All of our data is now in the spreadsheet and ready to use in a report.

Save the Excel file.

Power View

Now let’s use Power View to create some visualizations. In the main Excel workbook, click on the Insert tab in the ribbon and then under the Reports group (near the middle of the ribbon) click on Power View.

PVButton

This creates a new tab in your workbook for your report. Rename the tab to SandyEvacuation. In the Power View Fields task pane on the right, make sure you choose All and not Active so that every table in the model is visible. Uncheck any fields that have already been added to the report.

Where it says “Click here to add a title” enter a single space (we’ll create our own title in a different location later). In the Power View tab click on the Themes button in the Themes group. Choose the Solstice theme (7 from the end). This theme shows up well on the dark, stormy background we will add later and it has colors that provide good indicators for our numbers.

Now let’s add some data to the report. In the Power View Fields task pane, expand the flashflood table and add a checkbox to State then to County. This will create a new table on the left. Add HR1, HR3, HR6, HR12, and HR24 from the flashflood table in that order (you can change the order later at the bottom of the task pane). Add Latitude and Longitude from flashflood. Expand census and click on PctBuckets. Now we have the data we need to build a map-based visualization that will show rainfall across a geographic area.

Note: If you get a message at any point asking you to “enable content”, click on the button in the message to do so.

Map

In the Power View Fields task pane go to the Fields section at the bottom. Click on the arrow to the right of Latitude and Longitude and change both from sum to average.

PVCensusFields

With the focus on the table in the main window, click on the Design tab then Map within the Switch Visualization group on the left. This creates one map per state and changes the choices available in the task pane. Change the following fields at the bottom of the screen to get a map that allows us to see the path of Hurricane Sandy as it crosses New England.

TILE BY = [blank]
SIZE = HR6 (drag it from the field list to replace HR1, set it to minimum)
LOCATIONS = Remove the existing field and drag down the Geography Hierarchy from flashflood
LONGITUDE = Longitude
LATITUDE = Latitude
COLOR = PctBuckets
VERTICAL MULTIPLES = [remove field so it’s blank]
HORIZONTAL MULTIPLES = [blank]

PVCensusMapFields

With the focus still on the map, click on the LAYOUT tab in the ribbon. Set the following values for the buttons in the Labels group, they will apply to this particular visualization (the map):

Title = none
Legend = Show Legend at Right
Data Labels = Center
Map Background = Road Map Background

PVMapViz

Go to the Power View tab, make sure the focus is on the map. Click on Map in the Filters section to the right of the report. Click on the icon to the far right of State. This adds another icon (a right arrow) for “Advanced filter mode”, click on that. Choose “is not blank” from the first drop down. Click on “apply filter” at the bottom. Repeat for County. This eliminates irrelevant data with no matching rows.

PVCensusFilter

Make it Pretty

Go to the Power View tab so we can add some text. In the Insert group click on Text Box. It will insert a box somewhere on the page, paste in “Evacuation Targets – % of Population Over 69 by County” (without the quotes). Size the box so that it is a long rectangle with all the text on one line, then move the box above the map. Repeat for a text box that says “Hurricane Sandy 2012” and move that above the previous text box. Make the text size of the last box = 24 in the TEXT tab (resize the box if needed).

Hint: To move an object on the report, hover over the object until a hand appears then drag the object.

PVTitles

Download some Hurricane Sandy pictures. In the Power View tab go to the Background Image section. Choose Set Image and insert one of your hurricane images. Set Transparency to 70% or whatever looks best without overwhelming the report. You may need to change the Image Position to stretch. In the Themes group of the Power View tab set Background to “Light1 Center Gradient” (gray fade in to center). Move the two titles and the map as far into the left corner as you can to leave room for the other objects. If you like, reduce the size of the map a bit (hover near a corner until you get a double-headed arrow) add a couple more pictures on the right side (Insert group, Picture button).

PVBackgroundMap

Save the Excel file.

Slicers

Now let’s create some slicers and add another chart to make it more visually appealing and useful.

Click in the empty area under the map, we are going to create a new object. Click on flashflood.State in the task pane to add that column to the new table. Click on the scroll bar on the right side of the table to make sure the focus is on the table. Click on Slicer in the Design tab. Stay in the Design tab and under the Text section click on the “A with a down arrow” button to reduce the font by one size. If necessary, do some minor resizing of the table so it doesn’t fall off the page or cut off a state name in the middle but still shows a handful of rows. Expand Filters on the right side of the screen and drag flashflood.State to the filters section. Choose the Advanced button, “is not blank”, and “apply filter”. Repeat these steps to create a separate slicer for County. Now when you click on a particular county, for example, in the slicer everything on the report will be filtered to that one county.

PVSlicers

Save the Excel file.

Rainfall Chart

Click in the empty space to the right of the new County slicer. Add ShortState.ShortState and 5 separate flashflood columns, one for each time increment collected: HR1, HR3, HR6, HR12, HR24. In the Design tab choose “Other Chart” then Line. In the “Power View Fields” task pane make sure TILE BY is empty, VALUES for HR1-24 = AVERAGE, and AXIS = ShortState.

PVRainfallFields

Reduce the font size by 1 in the Text group of the Design tab then resize to make sure all values from CT to WV show up in the chart (stretch it to the right). In the Power View tab click on Insert.Text Box, move the box to be above the new chart, and type in the text “Average Rainfall by State”.

Make sure the focus is on the rainfall chart and go to the LAYOUT tab:

Title = None
Legend = Show Legend at Right
Data Labels = Above

PVHurricaneSandy

Save the Excel file.

That’s the completed report using the same steps as the report I demo’d at PASS Summit 2012 (download that report from the bottom of the blog). Try changing the SIZE field to HR1 (rainfall projection for the next hour), HR6, etc. Click through on the state map values to see the more detailed county level values. Click on a slicer value to see how that affects the map.

What you see in this report is a combination of SQL Server (structured) data and Hive (originally unstructured) data all mashed up into a single report so the end user doesn’t even have to know about the data sources. You can add more types of data, change the granularity of the data, or change how you visualize the data as you explore. The Hive data in particular is very flexible, you can change the structure you’ve imposed without reloading the data and add new data sources without lots of pre-planning. Using BI client tools to access all the additional rich data companies are starting to archive and mine in Big Data technologies such as HDInsight is very powerful and I hope you are starting to see the possibilities of how to use Hive in your own work.

I hope you’ve enjoyed this small bite of big data! Look for more blog posts soon on the samples and other activities.

Note: the CTP and TAP programs are available for a limited time. Details of the usage and the availability of the CTP may change rapidly

DemoSandy2012.zip