Collecting data via Capacity Agents
Gateway Server and Capacity Agents enable you to collect data from the required servers, nodes, or hypervisors and send the collected data to the Application Server for managing and estimating the capacity of resources that are used in your environment.
Install Capacity Agents on the systems from which you want to collect data. You must install the Gateway Server on the supported platform and configure with the Application Server (that hosts the TrueSight Capacity Optimization console) to perform the following tasks:
- Manage the collection of data from the required servers.
- Process the collected data (Convert it to a format that can be used for capacity planning in the Application Server).
- Send the processed data to the Application Server.
Additionally, you can configure the Gateway Server to aggregate and store the process-level data from the Capacity Agents. This process-level data can be accessed from the Process Profile page in the Data Explorer view for AIX, KVM and Solaris in the TrueSight console.
This topic provides the following information:
Configure the Gateway Server and Capacity Agents to collect data
To configure the Gateway Server and Capacity Agents to collect data
- In the TrueSight Capacity Optimization console, add Gateway Server in the Gateway Manager. For more information, see Adding, removing, and inspecting Gateway Server details.
- For the new Gateway Server, enable Auto-create ETL and schedule the ETL task. This creates an ETL task chain that is used for controlling all ETLs for the selected Gateway Server. For more information, see Automatically creating ETLs.
- Add the agent list to the Gateway Server to automatically create policy files and Manager runs. The Gateway Server splits the list of servers into multiple domain files. Each domain file is then split into multiple Manager Script files that contain the configuration required for creating Manager runs. For more information, see Managing the Gateway Server agent list.
After you submit the agent list, the data collection and processing is automated, and system creates an ETL task for each group of five Manager runs.
- (Optional): Modify the default configuration for the new Manager runs. For more information, see Creating new Manager runs. You can also modify the default workload configuration for the Manager runs. For more information, see Manual data collection using a Windows Gateway Server.
- Schedule the new Manager runs. If there are multiple Manager runs created from an agent list, you can schedule the Master Manager run. By default, the schedule that is configured for the Master Manager run is used for all its associated Manager runs. For the required Manager runs, you can set a schedule that is different from the schedule set for the Master Manager run. For more information, see Scheduling new Manager runs.
Manager runs initiate the data collection on the required server and store the collected data in the repository.
- (Optional): View the scheduled Manager runs. For more information, see Viewing scheduled Manager runs.
- Check the status of Manager runs and recover the failed Manager run (if any). For more information, see Managing the Manager runs.
- View the Gateway reports to view the details related to the status, errors and exceptions. For more information, see Accessing Gateway reports.
After the data is collected, the Gateway Server components (Analyze, Predict) convert the data to the vis file format and the Auto-create ETLs then automatically send the data to the Application Server.
Configure the Gateway Server to aggregate process-level data
Before you begin
Configure the Gateway Server and Capacity Agents to collect data. For more information, see Configure the Gateway Server and Capacity Agents to collect data.
To configure the Gateway Server to aggregate process-level data
- In the TrueSight Capacity Optimization console, select the required Gateway Server.
- Schedule process rollup for aggregating and storing process-level data from AIX, KVM and Solaris servers. For more information, see Aggregating process-level data for views.
- In the TrueSight console, use the Process Profile page of the Data Explorer views to view the process-level data. For more information, see Using views to analyze infrastructure and virtual data.
The data collected by Capacity Agents (in the UDR format) is periodically transferred to the Gateway Server where it is automatically processed into hourly intervals and saved in text files called VIS files. This processed data is then sent to the Application Server. Review the following guidelines about retention period for VIS files and disk space requirements.
- Keep about three months' worth of vis files. These are useful in case you need to "recover" data. For more information, see Recovery section.
- For UDR and vis files, it is recommended to keep three months worth of data, of which the oldest two months could be kept compressed to save space.
- Review UDR disk space requirements
- Review Vis files disk requirements
- Each manager run should have no more than 4 domains (one domain is one vis file)
- One domain should have no more than 25 nodes (managed systems)
Platform classifications for Capacity Agents
|Platform classification||TrueSight Capacity Optimization system types for compute|
|A machine running a PowerVM hypervisor||Virtual Host - AIX PowerVM|
|AIX in a PowerVM LPAR/SPLPAR||Virtual Machine - AIX LPAR; Virtual Machine - AIX Shared Processor LPAR|
|VIO Server in PowerVM||Virtual Machine - AIX VIO LPAR|
|A workload partition in AIX||Workload Partition - AIX WPAR|
|Solaris in any LDOM/Zone/DSD||Virtual Host - Solaris; Virtual Machine - Solaris LDOM; Virtual Machine - Solaris Zone; Virtual Machine - Solaris Dynamic System Domain|
|HP Integrity||Virtual Host - HP Integrity; Virtual Machine - HP Integrity|
|HP-UX virtualization||Virtual Host - HP nPar/vPar; HP nPartition; HP vPartition|
|Xen hypervisor and Oracle-VM||Virtual Host - Xen; Virtual Machine - Xen|
|Microsoft Hyper-V||Virtual Host - HyperV; Virtual Machine - HyperV|
|KVM||Virtual Host - KVM; Virtual Machine - KVM|
|Any operating system running directly on a "standalone" x86 based machine||Generic|
|Any operating system running on an x86 based hypervisor||Virtual Node - HyperV; Virtual Node - KVM; Virtual Node - VMware; Virtual Node - Xen|