Performance and scalability recommendations for integrations with CMDB
This topic presents recommendations and considerations for TrueSight Infrastructure Management integration with CMDB.
Timing for event and impact management operations
This topic presents timing approximations for operations in an environment that includes event and impact management:
The following environment is used as a sample for the timing estimates.
Metric | Amount or value |
---|---|
Number of CIs | 50 K |
Number of service models | 3 |
Number of levels | 6 |
Number of events per day | 400 K |
Maximum number of CIs in a service model | 20 K |
Number of impact events out of 400 K | 312 K |
Publish times
The following table shows the amount of time it takes to publish a new service model in each environment
Environment | Publish time |
---|---|
Small (5 K CIs and relationships) | 32 minutes |
Medium (10 K CIs and relationships) | 39 minutes |
Large (20 K CIs and relationships) | 1 hour |
The following table shows the amount of time it takes to publish a service model in which 1,000 CIs and relationships are added, updated, or deleted. Times are measured on a large environment.
Action | Publish time |
---|---|
Add | 16 minutes |
Update | 19 minutes |
Delete | 30 minutes |
Sizing chart for BMC Cloud Life Cycle Management
The following table can help you determine the number of required TrueSight Infrastructure Management Central Servers, TrueSight Infrastructure Management Child Servers, Data Collection Hosts, and BMC Server Automation Servers. The table also shows the maximum number of devices, monitor instances, and attributes for a proof of concept (POC) and large environments in BMC Cloud Lifecycle Management.
About a POC deployment
In a POC deployment, all of the components for BMC Cloud Life Cycle Management, such as the TrueSight Infrastructure Management Central Server, Child Server, and Data Collection Host, are on the same computer.
Characteristic | POC | Large environment |
---|---|---|
Number of monitored devices | 500 | 50 K |
Number of monitor instances | 4.5 K | 250 K |
Number of attributes | 8 K | 500 K |
Duration RAW data retention | 8 days | 8 days |
Duration RATE retention | 3 months | 3 months |
Adapter type | BMC PATROL | BMC PATROL |
Polling frequency in minutes (stats, auto-sync) | 5.5 | 5.5 |
Required number of TrueSight Infrastructure Management Central Servers | 1 | 1 |
Required number of TrueSight Infrastructure Management Child Servers | 1 | 5 |
Required number of TrueSight Infrastructure Management Data Collection Hosts | 1 | 5 |
Required number of TrueSight Network and Server Automation servers | N/A | 1 |
Required number of computers | 1 | 11 |
Required number of processors | 2 | 32 |
Required memory in GB | 8 | 128 |
Timing for events or incidents that can be sent
This topic provides information about the number of events or incidents that can be sent within a specified time frame.
Use cases | Events per minute | Duration in hours |
---|---|---|
L1 Event to incident creation | 60 | 48 |
L2 CIs impacting with incident information | 50 | 48 |
L3 Business service impact model | 35 events on TrueSight Infrastructure Management 5 events on the PATROL Agent | 48 |
Tuning parameters
In the <installed_location>\pw\integrations\ibrsd\conf\IBRSD.conf file, configure the #Executor threadpool
size as shown below:
ThreadPool.initialSize=10
ThreadPool.maxSize=10
Comments