Impact of BMC Server Automation use cases on database size
This topic describes some typical
use cases and how they could impact the reports data warehouse size.The
database size requirements and reports data warehouse performance are impacted by the way that you use . Several use factors influence the database size. Some factors are domain-specific; others apply to all domains.The following size factors apply to all product environments, regardless of the domains that you use:
- The number of databases (sites) from which data is transferred to the reports data warehouse is the foremost sizing factor.
Transferring data from additional sites can result in large increase in data warehouse volume. Therefore, it is important to carefully plan the number of sites included in the initial data transfer and your subsequent site additions for data transfer.
You should use the sizing tools provided with the product to determine database size requirements per site and plan your database storage, memory requirements, and CPU capacity requirements. For more information about sizing tools, see Sizing-tools. - The number of target servers for which you collect data (and subsequently transfer data to the data warehouse) impact your database size requirements.
The data volume resulting from a job is impacted by the number of servers for which data is collected. The volume resulting from a job run depends on the type of job. For example, Snapshot, Audit, and Compliance Jobs generate more data than other jobs.
If you are collecting data for a large numbers of servers (tens of thousands), the data generated in the job runs and then transferred to the data warehouse during each ETL run can be several GB in size.
The frequency at which you run jobs impacts your database sizing requirements and performance. Try to find a balance between your reporting needs and the volume of data handled by the data warehouse and run jobs and transfer data only as needed. For example, if you need to generate compliance reports once a week, you might not need to run Compliance Jobs every day and doing so might be an unnecessary burden on your system.
In addition to the sizing issues described above, which are not specific to reporting domains, the following table lists some domain-specific use cases that influence database sizing.
Domain-specific use cases influencing database sizing
Domain | Use case |
|---|---|
Inventory, Change Tracking | Data for these domains is generated in BMC Server Automation using Snapshot Jobs. The data generated by Snapshot Jobs primarily depends on the number of template parts (or assets), such as a service, a file, or a registry entry. The first run of a Snapshot Job generates lot of data, because it creates a baseline of the states of the current assets. For example, if a Snapshot Job is using a folder with 10,000 files on all servers, the baseline causes 10000 incremental ETL runs, and there is a sudden increase in data and entries per target server. The amount of data generated by a Snapshot Job using an Extended Object depends on the kind of output generated by the Network Shell script. It can be a large amount of data. Size requirements are further increased, because this data needs post-processing in BMC BladeLogic Decision Support for Server Automation for use in reports. The post-processing uses a lot of resources. It is important to carefully design Extended Objects to include the minimum amount of information that is required to satisfy your reporting needs. When ETL is executed for the first time, the warehouse is only populated with the latest baseline and its delta. This improves ETL performance and minimizes the use of warehouse space. |
Compliance | The amount of data generated by Compliance Jobs is directly proportional to the number of rules that are included. A Compliance Job generates the same number of rows per execution (unlike Snapshot Jobs). To control data volume, it is important to include only the rules needed for reporting purposes. |
Patch | The amount of data generated by Patch Jobs is directly proportional to the number of patches that are included. A Patch Job generates the same number of rows per execution (unlike Snapshot Jobs). To control data volume, it is important to include only the patches needed for reporting purposes. |
Job Activity (Job Run) | This domain includes a variety of event data logged in the BMC Server Automation data base by the BMC Server Automation Application Server. There can be several hundred events per job run, resulting in large amounts of data volume. For this reason, the default ETL runs do not transfer this data to the data warehouse. If you decide to transfer this data to the data warehouse, ensure that it is necessary for your business reporting needs and that your system is set up to handle the large sizing and processing requirements. |
Server Usage (agent log) | This domain includes data from logs files generated by BMC Server Automation agents on each managed server. In some cases, data volume can amount to several hundred GB per day. For this reason, the default ETL runs do not transfer this data to the data warehouse. If you decide to transfer this data to the data warehouse, ensure that it is necessary for your business reporting needs and that your system is set up to handle the large sizing and processing requirements. |
For more information about ETL runs, see Running-ETL.