3.11 Using Performance Data Import to Import Generic Data into PlateSpin Recon

PlateSpin has developed a robust Performance Data Import connector for PlateSpin Recon that allows users to import collected performance data from many third-party tools on the market into PlateSpin Recon to take advantage of its charting, reporting, and consolidation planning features.

When collected data has been moved from its proprietary location and format into a set of tables in the PlateSpin Recon monitoring database, PlateSpin Recon can import the data and make use of it like any other performance data. You can also set up a scheduled or recurring import, so the PlateSpin Recon analysis does not have to be a one-time thing.

Almost all performance data gathering tools use a different method and format for storing the gathered data. To enable PlateSpin Recon to successfully import any data, regardless of its original format, you need to use an Extract, Transform, and Load (ETL) tool. This type of tool takes the data, reshapes it, and then loads it into the tables where PlateSpin Recon expects to find it. See Section 3.11.2, Preparing Data for Import into PlateSpin Recon.

The workloads in the data being imported need to be of the type that PlateSpin Recon can inventory under normal circumstances. All necessary ports between PlateSpin Recon and the workload being imported must be open.

As with other remote data connectors, data to be imported should contain one or more data points for every hour. Also, to use Consolidation Planning, specific counters need to exist in the data. See Section 3.9.1, Required Performance Counters.

3.11.1 Performance Data Import Tables Data Structure

There are four tables used for importing of generic performance data:

ImportComputer table: Contains information about computers that are to be imported into PlateSpin Recon.
ImportCounter table: Contains information about all counters that were collected for any of the computers in the ImportComputer table.
ImportCounterInstance table: Contains information about named instances of each counter for each computer. The combination of Computer, Counter and Instance Name must be unique.
ImportData table: Contains monitoring data to be imported.

Each table contains one ID field, which acts as a primary key and is automatically generated by the database server when a new row is inserted into the table.

Table 3-8 ImportComputer Table

Column Name

Type

Description

Integer

Generated automatically.

Name

varchar(255)

Network name of computer.

This is the address or name of the computer. It allows PlateSpin Recon to address the computer over the network.

Table 3-9 ImportCounter Table

Column Name

Type

Description

Integer

Generated automatically.

Name

varchar(255)

Unique name of the counter.

This is the name of the counter as it will be shown in PlateSpin Recon charts. Some predefined counter names exist. See Section 3.9.1, Required Performance Counters.

Table 3-10 ImportCounterInstance Table

Column Name	Type	Description
ID	Integer	Generated automatically.
Computer	Integer	Reference to the ImportComputer table.
Counter	Integer	Reference to the ImportCounter table.
InstanceName	varchar(255)	Name of the given counter instance on the specific computer.

Table 3-11 ImportData Table

Column Name	Type	Description
ID	Integer	Generated automatically.
Instance	Integer	Tells PlateSpin Recon for which computer, counter and counter instance this data value was collected.
ObservationDate	DateTime	Time the data was collected.
ObservationValue	float	Value of the counter.

3.11.2 Preparing Data for Import into PlateSpin Recon

To migrate data from a third-party location and format requires an Extract, Transform and Load (ETL) tool. There are many such tools on the market:

Open Source: CloverETL, Enhydra Octopus, Pentaho Data Integration, and Talend Open Studio.

You can also write proprietary data manipulation scripts, such as SQL queries, although the specifics of these tasks are beyond the scope of this document. For third-party tools, see the documentation that accompanies those tools. For scripts, refer to your in-house reference material on the languages and requirements.

Commercial: Informatica PowerCenter, Inaplex Inaport, and Altova MapForce.

Regardless of the tool you choose, the data transformation workflows are the same:

Retrieve hostnames of the workloads you want to import and write them to the ImportComputer table
Write the names of the required counters to the ImportCounter table
Retrieve and write all performance data related to the above counters

To work with generic imported data:

Collect data in the required performance and utilization measurement categories. For more information, see Section 3.9.1, Required Performance Counters.
Extract, transform, and load the collected performance data into the PlateSpin Recon Performance Data Import tables.

For more information, see Section 3.11, Using Performance Data Import to Import Generic Data into PlateSpin Recon, Section 3.11.1, Performance Data Import Tables Data Structure and Section 3.11.2, Preparing Data for Import into PlateSpin Recon.
Create a new Remote site for connecting to the Performance Data Import tables:
1. Right-click in the Data Center Explorer, click Sites, then click Add Site.
2. In the Add Site dialog box, specify a name and, optionally, a description.
3. In the Type drop-down, select Remote.
4. In the Connector Type drop-down list, select Performance Data Import.
  
  The Delete data from tables when done? check box is displayed.
5. To empty the Performance Data Import tables of the data loaded in Step 2, select the Delete data from tables when done? check box.
  
  If you choose not to allow PlateSpin Recon to empty these tables, you must manually do this before performing another import.
6. Click OK
  
  The new site is added to Sites node in the Data Center Explorer.
Inventory the machines:
1. In the Data Center Explorer, right-click the remote site you created in Step 3, click Sites, then click Inventory Machines.
  
  The Inventory Machines in Site sitename dialog box is displayed.
2. Select the check boxes next to the machines you want to inventory.
3. Select the required credentials or select <New Credential> to add new credentials.
4. Click OK.
Synchronize site data:
1. In the Data Center Explorer, right-click the required site, click Sites, then click Synchronize Data.
2. Select the check boxes next to the required machines. To select all machines in a site, select the check box next to the site.
3. Select required data range, scheduling, and recurrence options.
  
  To perform a recurring import, you need to also set up a matching recurring Extract, Transform, and Load process to refresh the data in the PlateSpin Recon Performance Data Import tables. For detailed information, see your software vendor documentation.
4. Click OK.