Deployment of Hadoop-based Services on the Windows Azure Portal

Deployment of Hadoop-based Services on the Windows Azure Portal

This topic describes using the Hadoop on Windows Azure Portal to provision a new Apache Hadoop cluster. Clusters provisioned on the portal are temporary and have an expiration. These clusters are provisioned to run jobs processing input data that may be on the cluster or located elsewhere. For example, the data could reside in a Windows Azure acount.

The advantage of using a temporary cluster is that there is no cost to maintain hardware used for parallel processing of map reduce jobs once the job is completed. You use the cluster then release it or allow it to expire. Apache Hadoop deployment solutions are also available for deploying Apache Hadoop to a Windows Azure account or on-premise hardware that you manage yourself.

The Windows Azure Map/Reduce Service (or the Hadoop-based service on Azure) is by invitation only during the Community Technical Preview (CTP).  The purpose of this CTP is for you to test Hadoop-based service on Azure, become more familiar with it, and provide feedback.  The process is:

 

1)     Fill out the linked Hadoop-based service on Azure CTP Connect Survey providing information about yourself, business domain, and technical scenario.  You will need a LiveID (e.g. Hotmail.com, Live.com) and a filled out Connect profile.

  • Click the link to find out more about the Microsoft Connect program.

 

2)     Once you receive an invite code, go to HadoopOnAzure.com and click Sign In.

 

3)     If it is your first time, you will be asked to allow HadoopOnAzure.com access to your LiveID profile information.  Please click, Yes.

4)     Under the “Got a code?” dialog, enter in your invite code, and click Enroll.

 

5)     For your new account, your first task will be to create a Hadoop on Azure cluster.  Your options are:

  • DNS Name: Choose a DNS Name that you would like to name your cluster
  • Cluster Size: Choose what size of cluster you would like to test against
  • Cluster Login: Provide username and password information so you can log into your cluster.

Once done, click Request cluster (right side bar, under the green bar)

 

6)      At this point, the Hadoop on Azure service is creating the new cluster.  This will take a few to tens of minutes to create depending on the number of nodes and the number of clusters being created at the time.

 

7)     Once the cluster is setup, your portal page will be similar to the one below.   Now you are good to go!

 

 

See Also

 

Leave a Comment
  • Please add 8 and 8 and type the answer here:
  • Post
Wiki - Revision Comment List(Revision Comment)
Sort by: Published Date | Most Recent | Most Useful
Comments
  • BradSevertson edited Revision 10. Comment: got rid of dead links to onprem and old win Azure deployment

  • BradSevertson edited Revision 9. Comment: got rid of rference to FTP option

  • BradSevertson edited Revision 8. Comment: Got rid of the EMR language: Elastic MapReduce Portal -> Hadoop on Windows Azure Portal

  • Ed Price - MSFT edited Revision 4. Comment: White space, formatting, added tags

Page 1 of 1 (4 items)
Wikis - Comment List
Sort by: Published Date | Most Recent | Most Useful
Posting comments is temporarily disabled until 10:00am PST on Saturday, December 14th. Thank you for your patience.
Comments
  • Ed Price - MSFT edited Revision 4. Comment: White space, formatting, added tags

  • BradSevertson edited Revision 8. Comment: Got rid of the EMR language: Elastic MapReduce Portal -> Hadoop on Windows Azure Portal

  • BradSevertson edited Revision 9. Comment: got rid of rference to FTP option

  • BradSevertson edited Revision 10. Comment: got rid of dead links to onprem and old win Azure deployment

  • These screen shots didn't exactly match what I saw during cluster deployment -- I didn't get the cluster size option and the allocation in progress screen was also different.

  • Thanks Jose. These screen shots are from an earlier version and need updating. We missed these ones in the last round of edits evidently. I will fix them. Cheers, Brad Severtson

  • Insert image functionality is not working. Will try later.

  • Nice article

Page 1 of 1 (8 items)