Applies to:
| TBMAs |
Objective: | Demonstrate how to configure Service Desk Tickets to ensure performance and scalability |
Pre-Requisites: | None |
Author: | Paul Sheldon and Greg Nutt |
Revision | Author | Description of Changes |
1.1 | Jeff Stark | Updated Template |
1.2 | Jesse Sharp | edit to make CFD |
| | |
| | |
| | |
Problem
The Tickets object is often backed to an extreme level of granularity without thought or consideration of whether or not you need reporting down to the level of the individual Ticket ID. The standard reports will usually group this data down automatically. So why add granularity in the object when the only impact is on performance?
Resolution(s)
Adjust the Object Identifier
The standard Object Identifier for the Tickets object is set to the individual Ticket ID. However, as described above, this level of detail is not required for reporting and is often too granular to even add much information. Instead, the Tickets object can be backed by the combination of a few fields to still provide the granularity necessary for standard reporting but reduce the size of the allocation tables unnecessarily affecting performance: Application ID, Priority/Severity, Type, Category, Impact, and Location.
When adjusting the Object Identifier on a sample data set we can see a significant drop in the granularity of our object, reducing the row count from 48,303 rows of data to 5,851 rows of data – a reduction of about 85%.
Enhanced Reporting – OPTIONAL
If you require a view into the Ticket ID level detail, then a second modeled table can be added on to the Tickets object for reporting purposes.
This Tickets Detail object can be left at the full granularity of the provided Tickets data, but is kept out of the path of the standard flow of the Cost Model. This allows a drill into, but not through this detail.
NOTE: Utilizing this methodology will require the customization of standard reports and/or the configuration of new custom reports. For more information on modeling and reporting within Apptio, please check on TBM Connect ()
Remember!
Remember that the purpose of the information (data) modeled within Apptio is to populate reports that help us make decisions. If a data element can’t help us make a decision or allocate costs, then it’s not a good candidate for inclusion in an identifier.
{Various} is not a four-letter word, and seeing a {Various} isn’t necessarily a sign of bad configuration! In fact, if you eliminated details that weren’t being used for analysis and improved performance, it can actually be good configuration.
When making exceptions to standard granularity recommendations or adding new objects and reports, pursue the least invasive options first.
Lastly, keep in mind that Cost Transparency is not an accounting system. Most analysts who could make use of extremely granular transaction-level data already have access to that information. The level of detail we need for cost modeling, allocations, and strategic business decisions is often much less! Following the above suggestions and best practices can keep your finance layer, and calculation times, as efficient as possible.