Administrators can monitor, launch, and inspect aspects of their Big SQL service through Ambari, an open source Web-based tool for managing Hadoop clusters.¬† In this section, you will learn how to
- Launch Ambari.
- Inspect the overall status of your cluster.
- Inspect the configuration and status of your Big SQL service.
- Identify the software repository (build source) of your Big SQL service.
Allow 30 minutes to complete this section. To work through these exercises, you must have a working BigInsights environment, as described in the initial lab in this series.
Please post questions or comments about this lab to the forum on Hadoop Dev at https://developer.ibm.com/answers?community=hadoop.
In this exercise, you will launch Apache Ambari and verify that a minimal set of services are running so that you can begin working with Big SQL. You will also learn how to stop and start a service.
__1. ¬†¬†¬†¬†¬†¬†¬†¬† Launch a Web browser.
__2. ¬†¬†¬†¬†¬†¬†¬†¬† Enter the URL for your Ambari service, which was configured at installation.¬† For example, if the host running your Ambari server is myhost.ibm.com and the server was installed at its default port of 8080, you would enter
__3. ¬†¬†¬†¬†¬†¬†¬†¬† When prompted, enter the Ambari administrator ID and password.¬† (By default, this is admin/admin).
__4. ¬†¬†¬†¬†¬†¬†¬†¬† Verify that the Ambari console appears similar to this:
__5. ¬†¬†¬†¬†¬†¬†¬†¬† If necessary, click the Dashboard tab at the top of the screen and inspect the overall status of services that have been installed.¬† The previous screen capture was taken from a system in which all open source components provided in the IBM Open Platform for Apache Hadoop had been started.¬† The Big SQL service was also started.
__6. ¬†¬†¬†¬†¬†¬†¬†¬† Click on the Flume service in the panel at left.
__7. ¬†¬†¬†¬†¬†¬†¬†¬† Note that detailed information about the selected service appears.
__8. ¬†¬†¬†¬†¬†¬†¬†¬† Since Flume isn’t required for this lab, stop the service.¬† To do so, expand the Services Action drop-down menu in the upper right corner and select Stop.
__9. ¬†¬†¬†¬†¬†¬†¬†¬† When prompted, click Confirm Stop.
__10. ¬†¬†¬†¬†¬† Monitor the operation’s status in the pop-window that appears.
__11. ¬†¬†¬†¬†¬† When the operation completes, click the OK button at the bottom of the pop-window to close it.
__12. ¬†¬†¬†¬†¬† Return to the Ambari display and verify that the Flume service has stopped, as indicated by the red triangle next to the service.
__13. ¬†¬†¬†¬†¬† Optionally, start the Flume service.¬† From the Services Action menu, select Start.¬† Click OK when prompted and allow the process to complete.¬† Click OK again.
__14. ¬†¬†¬†¬†¬† Confirm that the following services are operational on your cluster before continuing with the remaining exercises:¬† HDFS, MapReduce2, Nagios, Ganglia, Hive, Knox, BigInsights – Big SQL.
Let’s explore the configuration of your Big SQL service, including the nodes on which various Big SQL artifacts have been installed.¬† This is important for subsequent exercises, as you need to know where Big SQL client software resides and where the Big SQL Head Node resides.¬† Big SQL client software includes JSqsh, a command-line interface that you’ll use in a subsequent lesson, and SFTP, which you’ll use later to LOAD data from a remote server into a Big SQL table.¬† To connect to Big SQL and issue commands or queries, you need to specify the JDBC URL of the Big SQL Head Node.
__1. ¬†¬†¬†¬†¬†¬†¬†¬† In Ambari, click on the BigInsights – Big SQL service to display details about it.
__2. ¬†¬†¬†¬†¬†¬†¬†¬† Examine the overview information presented in the Summary tab.¬† In the previous screen capture, you’ll note that there are two Big SQL workers installed and running.¬† Furthermore, JSqsh and SFTP are installed on 3 nodes.
__3. ¬†¬†¬†¬†¬†¬†¬†¬† Click on the Hosts tab towards the upper right of your screen.¬† A summary of the nodes in your cluster is displayed.¬† The image below was taken from a 4-node cluster.
__4. ¬†¬†¬†¬†¬†¬†¬†¬† Expand the components information for each node to inspect the installed services.¬† In particular, note which node contains the Big SQL Head node, which node(s) contain the Big SQL Workers, and which node(s) contain the Big SQL clients service (JSqsh).
__5. ¬†¬†¬†¬†¬†¬†¬†¬† Optionally, explore the software repository associated with your Big SQL installation.¬† In the upper right corner, click Admin > Repositories. Scroll to the bottom of the displayed page and note the repositories associated with the software you have installed on your cluster.¬† This information indicates the build levels for your core services (IBM Open Platform for Apache Hadoop), core utilities, and IBM value packs.