(1) There are three different sizing that might be requested:
(A) A sizing for a new environment
(B) A re-sizing of an existing environment
(C) A planned growth resizing of an existing environment
(A) For a sizing of a new environment you will need to fill out the spreadsheet enclosed in this knowledge article based upon the types of servers in your environment and open a ticket to TrueSight Capacity Optimization (TSCO) support including the spreadsheet for evaluation and validation.
(B) For a re-sizing of an existing environment (possibly one that wasn't initially sized correctly or is performing poorly), create a sizing report (https://bmcsites.force.com/casemgmt/sc_KnowledgeArticle?sfdcid=000152366) and a data flow report (https://bmcsites.force.com/casemgmt/sc_KnowledgeArticle?sfdcid=000025194). Also, provide the current deployment description, how many Application Servers and ETL Engines, their hardware configuration and the database configuration. The TSCO sizing team can then use that information to calculate the recommended hardware environment based upon the data currently being imported into CO.
(C) For a planned growth implement the attached both the reports listed in previous point and also you need to populate the sizing spreadsheet to indicate the expected growth rate or quantity from your current sizing.
The minimum hardware requirements are documented in the TSCO product documentation, for adequate sizing raise a case with support.
(2) To provide a sizing recommendation it is necessary to fill out the spreadsheet enclosed in this knowledge article based upon the types of servers in your environment and open a ticket to TSCO support including the spreadsheet for evaluation and validation.
(3) Sections A and B of the spreadsheet are designed to determine which ETLs are to be implemented and what type of data is to be collected by each ETL. This section is critical for sizing because the ETL data load directly impacts the sizing of the TSCO database.
For example : If you will be implementing the Gateway Manager ETL, then you'll fill out section " A. Processed Rows for Managed Systems and Business Drivers" with the planned counts of systems being imported into TSCO
In section "B. Data Sources" you'll specify the name of the TSCO Gateway Server [formerly known as the BMC Performance Assurance (BPA)] connectors you need to create.
There are 9 different data collection options for different system types for the Gateway /Vis Parser ETL.
- Standalone Servers (any OS)
- VMware
- AIX VM/WPAR (includes AIX DLPARs/SLPARs)
- Solaris LDOM/Zone/DSD
- HP Integrity Virtual Machines and/or nPar, nPar/vPar
- Xen
- Microsoft Hyper-V
- KVM (Linux Kernel-based Virtual Machine)
- Virtual nodes (BPA collectors installed within a VMware guest instance)
It is also possible to decide what type of data you want to import from the Gateway Server for those systems, the options are:
- Performance Data (required)
- Detailed Disk/Network Data (do you want individual disks/NICs imported from TSCO Agent into the TSCO database)
- Workload Data
- Command Statistics (The Top 30 commands from the VIS file)
- User Statistics
Note that currently the TSCO Sizing Spreadsheet considers the Detailed Disk/Network Data, Command Statistics, and User Statistics options to be unselected in the Gateway Server vis parser ETL. If these options are selected it will be necessary to increase the '# metrics' that are being imported by the Gateway Server vis parser ETLs. Contact BMC Technical Support for additional information regarding these considerations.
Pay attention to the "C. Aging Policies" section and decide if you'll be picking different data retention periods (although the defaults are typically appropriate for most environments).
In section "D. Users & Scheduled Reports" you'll just want to note how many people you expect to be using TSCO (this is more for CPU sizing than database space sizing).
Finally, in section "E. Deployment Options" the spreadsheet is just looking for you to indicate what platforms you will be deploying TSCO on and whether you are looking to implement a High Availability (HA) environment (failover) or whether you will be implementing a standard environment.
(4) When the TSCO Sizing Team returns their sizing analysis they will provide recommendations for TSCO AS, TSCO EE, and Database CPU and Memory sizing. They will also provide the predicted Database size and the expected number of Database IOPS expected.
Do you plan to use the TSCO Gateway Server in the capacity planning environment? If so, will the Gateway Server be used for the Analyze and Investigate features? Make sure to include this information when providing sizing data for the environment in the details.
Additional Details
Section A. Processed Rows for Managed Systems and Business Drivers
Entity Type | Description |
Physical Standalone Servers | The number of standalone physical servers (any platform) for which system level performance data will be imported into BCO in your environment. |
- workloads | For physical systems imported from TSCO Gateway Server/BPA, this is the number of TSCO Gateway Server/BPA workloads that will be imported into an environment. Note that this is the number of workload entities so it a sum of workloads to be imported across all physical systems where the BPA ETL will be configured to import 'Workload Data' |
AIX Partitioned Servers | The number of AIX frames to be imported into TSCO in your environment |
- workloads | For AIX partitions imported from TSCO Gateway Server/BPA, this is the number of TSCO Gateway Server/BPA workloads that will be imported into an environment. Note that this is the number of workload entities so it a sum of workloads to be imported across all AIX partitions where the TSCO Gateway Server/BPA ETL will be configured to import 'Workload Data' |
- LPARs/WPARs | The total number of AIX partitions (DLPARs, SPLPARs) to be imported into TSCO. This is the number of partitions that reside within the frames listed in the 'AIX Partitioned Servers' above. |
Solaris Partitioned Servers | The number of Solaris DSD and global zones. For the TSCO Gateway Server/BPA ETL this is basically a count of the number of 'installed collectors' (since the collector is installed at the physical system level) that will be collecting data to be imported into BCO. |
- workloads | For Solaris partitioned servers imported from TSCO Gateway Server/BPA, this is the number of TSCO Gateway Server/BPA workloads that will be imported into an environment. Note that this is the number of workload entities so it a sum of workloads to be imported across all Solaris partitions where the TSCO Gateway Server/BPA ETL will be configured to import 'Workload Data' |
- Containers/Zones/Ldoms | The total number of Solaris partitions and zones that will be collecting data to be imported into BCO. For TSCO Gateway Server/BPA, for zones in particular although there is only one installed collector running at the global zone level the individual local zones will each be represented as a separate system entity in TSCO (similar to how they are represented in TSCO Gateway Server/BPA). |
HP-UX Partitioned Servers | The number of HP-UX partitioned servers (the physical server on which HP nPar, vPar, and IVM guests reside) that will be imported into TSCO. |
- workloads | For HP-UX partitioned servers imported from TSCO Gateway Server/BPA, this is the number of TSCO Gateway Server/BPA workloads that will be imported into an environment. Note that this is the number of workload entities so it a sum of workloads to be imported across all HP-UX partitions where the TSCO Gateway Server/BPA ETL will be configured to import 'Workload Data' |
- IVMs/nPars/vPars | The total number of HP-UX IVMs, nPar, and nPar/vPar partitions to be imported into TSCO. |
ESX Virtual Servers | The total number of VMware ESX hosts to be imported into TSCO via the vCenter Service Extractor or vCenter History Extractor ETLs. |
- guest machines | The total number of VMware guest machines that reside on the ESX hosts to be imported into TSCO. |
XenServer Virtual Servers | The total number of Xen servers. For Xen data imported into TSCO via the TSCO Gateway Server/BPA ETL this is the number of physical Xen 'Dom0' servers on which the TSCO Agent/BPA agent is installed. |
- guest machines | The total number of Xen guest domU instances that reside on the Xen Dom0 hosts to be imported into TSCO. |
HyperV Virtual Servers | The total number of physical Hyper-V hosts to be imported into TSCO. |
- guest machine | The total number of Hyper-V guest OS instances that reside on the Hyper-V hosts to be imported into TSCO. |
Virtual Node | The number of ESX guest machines on which a collector will be installed within the guest OS instance itself to collect performance data which will then be imported into BCO. This is typically related to VMware guest OSs running a local TSCO Agent/BPA collector that is reporting 'VIRTUAL' node data. |
- workloads | For virtual nodes imported from TSCO Gateway Server/BPA, this is the number of TSCO Gateway Server/BPA workloads that will be imported into an environment. Note that this is the number of workload entities so it a sum of workloads to be imported across all virtual nodes where the TSCO Gateway Server/BPA 'Virtual Nodes' ETL will be configured to import 'Workload Data' |
Database Server | The number of Database Instances that will be imported into TSCO. These entities are typically via the Oracle OEM ETL. |
Network | The number of network entities that will be imported into TSCO. These entities are typically imported by one of the TSCO network ETLs. |
Storage | The number of of Storage LUNs what will be imported into TSCO . These entities are typically imported by one of the TSCO Storage ETLs, such as the NetApp ETL. |
Business Drivers | The number of Business Drivers to be imported into TSCO. Business Drivers typically come from a 3rd party database and are often import into BCO via a CSV parser or database extractor ETL. These are things like 'Number of web page hits', application transaction counts, and so on. |
B. Data Sources
The goal of this section is to identify the number of ETLs that are likely to be implemented in an environment.
TSCO Gateway Server/BMC Performance Assurance (BPA)
For the TSCO Gateway Server/BPA vis parser ETL the number of ETLs configured really depends on the size of the environment and whether the environment is using AutoETL to create the TSCO Gateway Server ETLs. By default AutoETL will create one ETL for every 5 Manager runs. Using the default Manger run configuration in the Gateway Manager "Agent List" configuration each Manager run will process data for 100 computers.
Please contact Technical Support for any assistance required with this calculation. In general the TSCO sizing recommendations are based upon the quantity of data (in millions of rows of data) to be imported and the number of ETLs is not a significant factor. This section mostly is used to detect whether there may be other factors that need to be considered for the sizing.
TSCO Gateway Server ETL configurable data types
The TSCO Sizing Spreadsheet considers the Detailed Disk/Network Data, Command Statistics, and User Statistics options to be unselected in the TSCO Gateway Server ETL. If these options are selected it will be necessary to increase the '# metrics' that are being imported by the TSCO Gateway Server ETLs. Contact BMC Technical Support for additional information regarding these considerations. If you are importing data via TSCO Gateway Server without the Disk/Network, Command, or User data options selected leave the '# of metrics' as the default since that default is the number of performance metrics imported by the default TSCO Gateway Server ETL configuration.
But, in the TSCO Gateway Server ETL there are then 4 different configuration options that would modify the number of metrics being imported into TSCO. These are:
- Detailed Disk/Network Data (Do you want individual disks/NICs imported from the TSCO Gateway Server into TSCO?)
- Workload Data (Do you want performance data for TSCO Gateway Server workloads import into TSCO?)
- Command Statistics (Do you want the Top 30 commands from Visualizer import into TSCO?)
- User Statistics (Do you want the User Statistics [by user aggregated process resource consumption] imported into TSCO?)
By far the most common configurations are:
(A) None of the above 4 options selected. This imports just standard system level performance data (no individual disks, network interfaces, workloads, commands, or user data)
(B) For TSCO Gateway Server customers with well-defined workloads just the "Workload Data" option selected. If you are planning on importing TSCO Gateway Server workloads you'll want to estimate the total number of workloads to be imported into TSCO by entity type and fill in the " - workloads" row for each entity type. (This is already supported by the standard TSCO sizing spreadsheet).
If you want to do Detailed Disk/Network Data for your TSCO Agent computers contact TSCO Technical Support to assist with that sizing as it will be necessary to estimate the number of Network Interface Cards and Disk devices that will be reporting data into TSCO.
Very few people select the 'Command Statistics' or 'User Statistics' options because they have a significant impact on the volume of data that is imported into TSCO each day. But if those are planned to be enabled let TSCO Technical Support know if it would be for all TSCO Agent systems import into TSCO or just a subset and the TSCO sizing can be calculated based upon that additional data.
TSCO Aging Configuration
In TSCO the Aging Configuration can be specified with different retention periods for Short, Medium, and Long aging. The Sizing Spreadsheet only has you specify a single aging configuration for each summarization period. This should be the "Long" aging configuration planned for your TSCO deployment as the vast majority of data is stored, by default, at the Long aging level.
Check the Aging values in your Sizing Spreadsheet as the default values in the sizing spreadsheet might not match the default values for the TSCO version you will be deploying. For example, in TSCO 10.7 and later the default aging configuration values are:
Detail (Long): 1 month
Hour (Long): 6 months
Day (Long): 12 months
Month (Long): 36 months
Confirm that the specified Aging configurations in your sizing spreadsheet are appropriate for the environment.
Legacy ID:KA350197