When installing and configuring EcoSys, following the steps listed below help ensure your EcoSys application and the infrastructure on which it is running are best suited to deliver the most optimal performance for your data, user, and functional needs.
Analyze Data Volume, Shape of Data, and User Volume
Typically, concerns over system performance center around two issues: data volume and user load. It is important to understand the volume of data expected to be in the system. However, it is equally important to understand the shape of that data, meaning the ratio of different object counts within the functional design.
Some ratios to consider in determining the shape of the dataset are:
-
How many root cost objects?
-
How many levels of child cost objects?
-
How many tasks per cost object?
-
How many actuals/forecast transactions per cost object?
-
For each transaction, how many custom field values and category assignments?
-
How many versions, and how will transactional data grow during the closeout process?
Knowing the volume and the shape your production data is a critical first step in knowing how to tune the technical infrastructure as well as the application configuration.
Analyze LANA Definitions
LANA definitions refer to any of the views configured on top of the LANA engine. These definitions include spreadsheets, reports, forms, actions, formulas, and published web services.
The common perception that overall data volume stored in the database is what most affects load times is usually not true. The most important factor affecting load times is more frequently the LANA definitions themselves. The second step in knowing how to tune a system for high performance is to document clearly the types of LANA definitions that will be used and their row and column filters.
Sizing the Infrastructure for LANA
Infrastructure components (servers, network, storage, and so forth) play an important role in performance. After estimates are available for the volume of data, number of concurrent users, and how the data will be accessed (LANA definitions), the infrastructure resources required may be addressed. For more details on infrastructure sizing recommendations please see sections of this document labeled with the Scalability and Sizing icon (). You may to do do this step with the performance benchmark tests mentioned below.
Optimizing Functional Configuration
LANA definitions (spreadsheets, reports, actions, forms, and published web services) support many options and settings that control how they perform. It is recommended that teams involved in performance tuning and advanced configuration become familiar with these options and settings, and then refine the relevant LANA definitions accordingly. For information about these options, see Configuration Optimizations for Spreadsheets, Reports, and Actions.
EcoSys also includes detailed performance instrumentation that assists with identifying bottlenecks. As the process of optimizing the configuration of LANA definitions goes on, these metrics can provide a feedback loop for refining the configuration for best performance. For more information about how to evaluate the performance of your LANA definitions, see Identifying Bottlenecks and Interpreting Performance Metrics.
Optimizing System Configuration
In addition to optimizing individual LANA definitions, EcoSys includes a set of options at the server level that relate to performance. These server-level settings should be adjusted to fit the target shape and volume of data and the user load. For more information about these settings, see EcoSys Server Settings.
Perform Data Load and Concurrent User Benchmark Testing
Measuring real-world performance is key in validating that the configuration and infrastructure meets the performance expectations for a given instance. We recommend you load the system with data volumes based on several years’ projected growth and with the expected concurrent user load.
The EcoSys technical support team can provide guidance on planning and executing performance tests, including recommending automation tools and analyzing results.