Performance Lab Metrics
Eagle Investment Systems conducted a focused testing exercise in the Eagle Performance Lab for generic inbound streams as an initial phase of performance testing around data interfaces. There are a multitude of generic inbound message streams available as part of Eagle's product suite.
This exercise focused on several commonly used streams:
Securities – Eagle_default_in_star_smf
Prices – Eagle_default_in_star_price
Trades – Eagle_default_in_star_trades
Positions PACE upload – Eagle_default_in_star_ThirdParty2Pace
Disposal Lots PACE upload – Eagle_default_in_star_ThirdParty2Pace
Trades PACE upload – Eagle_default_in_star_ThirdParty2Pace
Cash PACE upload – Eagle_default_in_star_ThirdParty2Pace
Eagle Lab performance measurements are shown in the table below.
Stream | Function | Records/Second |
---|---|---|
Eagle_default_in_star_smf | Load Securities | 28 records/second |
Eagle_default_in_star_price  | Load Prices | 56 records/second |
Eagle_default_in_star_trades | Load Trades | 3 records/second |
Eagle_default_in_star_ThirdParty2Pace | Load Positions, Lots | 13 records/second |
Eagle_default_in_star_ThirdParty2Pace | Load Cash Activity | 23 records/second |
Eagle_default_in_star_ThirdParty2Pace | Load Custody files | 56 records/second |
This exercise demonstrated that some of the key generic inbound message streams had respectable performance throughput results, but there were others where the performance throughput results were not as strong as preferred.
By performing this exercise, Eagle established a throughput baseline for the above streams to measure improvements made going forward. During the testing exercise, we were able to identify several specific areas to enhance in future releases and help improve throughput.
Eagle is focused and committed to expanding the breadth of generic interfaces, inbound and outbound. Continuous testing enables Eagle to evaluate the loading mechanism and incorporate additional changes to the EagleML interface standards, offering clients additional choices for improved throughput.
For more information about EagleML, visit the EagleML wiki pages.