IBM PureData System for Analytics, Version 7.1

Deploying IBM Netezza Platform Development Software

IBM® Netezza® Platform Development Software 7.1 deployment package for vSphere 5.x is a single OVA file that is shipped as a zip archive file. You must unpack and then deploy the file using vSphere Client.

Before you begin

Unpack the zip file on the administrative workstation where vSphere Client is installed. The INPDS.ova file is created. It contains a template of VMware vApp with two virtual machines:

NPS® – This virtual machine works as NPS front-end host to which the user connects to manage database or process SQL queries. The machine has two virtual network interfaces. The first one is for external connections. The second one is for internal connectivity with SPUs.
SPU01 – This is a Snippet Processing Unit virtual machine that is managed by NPS virtual machine. The machine is connected with NPS by using its single network interface – it is not accessible for external world. The machine has four virtual hard disks that are used for storing user data in the data warehouse.

Important: Do not change the names of these virtual machines.

Procedure

Run vSphere Client and connect to vCenter Server.
From the File menu select Deploy OVF Template.
On the Source panel, use Browse to select the INPDS.ova file location, or type the path in the Deploy from a file or URL text field. Click Next.
On the OVF Template Details panel, verify product name and version: IBM Netezza Platform Development Software 7.1. Click Next.
On the Name and Location panel, you can change the vApp name. This name must be unique in the entire data center.
On Inventory Location panel, select data center into which IBM Netezza Platform Development Software is to be deployed. Click Next.
On the Host / Cluster panel, select the ESXi host or cluster that is the deployment target. Click Next.
On the Resource Pool panel, select the resource pool within which you want to deploy this template. Click Next.
On the Storage panel, select data store or storage cluster to deploy the virtual disk images to, and click Next.
Note: During deployment all images must go to the same data store. You can change this setting after the deployment is finished.
On the Disk Format panel, select the provisioning method.
Tip: Select Thick Provisioning Lazy Zeroed or Thick Provisioning Eager Zeroed only if all the virtual disks are to be stored in the same data store, and only if there is enough space to allocate. Otherwise, select Thin Provisioning.

Click Next.
On the Network Mapping panel:
1. Select External network - required for external communication to IBM Netezza Platform Development Software (using SSH, ODBC, JDBC, OLE DB, and so on).
2. Select Internal network - required for internal communication between NPS and SPUs, that is, isolated and private connectivity between all virtual machines from IBM Netezza Platform Development Software. Select network name or port group name that you created in Preparing the internal network, or contact vSphere administrator.
Click Next when done.

Note: At this point, the network settings of your choice are not verified automatically. For that reason, pay attention while defining the internal and external network to avoid potential network configuration problems. If these networks are defined incorrectly, either the network configuration, or the NPS installation step fails.
On the Ready to Complete panel, verify the deployment settings. Click Finish to start the deployment.

Results

The deployment progress dialog is displayed. The ova file size is ~ 3.5 GB, and the amount of time that is required for deployment depends on network throughput. Typically, it is between 10 minutes and 1 hour.