Redemptions
Concurrency limits
Redemptions are the orders generated by members when they redeem their available Currencies (e.g. the standard Points) for Rewards. Redemptions include one or more Redemption Items.
Redemptions in Fielo have gone through a rigorous stress test to help understand the upper concurrency limits (transactions being generated at the same time) that you may run into as the members in your program grow.
There are three ways to manage reward redemptions:
Each one of these has a different outcome in the performance test, which you can find below:
Salesforce Sandbox configuration
To understand our testing methods better, we have provided the details of our testing Sandbox configuration:
- Members: 500k registered in a program
- Levels: 3 levels
- Segments: 3 segments
- Storage: 30Gb of usage
Rewards with Unlimited Stock
Redemptions of unlimited stock reward can handle up to 1.6 million transactions per day, independent of how many rewards you may have in your program. In this scenario, we tested redemptions with 10 different rewards.
The following graphic shows the number of transactions (redemptions) per second. The main takeaway from this test is we can guarantee up to 100 transactions per second. The above 100 transactions may introduce errors.
Rewards Controlled by Quantity
Redemptions controlled by quantity can handle up to 1,200,000 transactions per day.
The following graph shows the number of transactions per second that it takes to execute the redemption of a reward controlled by quantity; in this scenario, we tested redemptions with 10 different rewards. Note the yellow line hovers around limit of 15 transactions per second.
Rewards Controlled by Item
Redemptions of rewards that are controlled by Item can handle up to 1,2 million transactions per day, independent of how many rewards you may have in your program. In this scenario, we tested redemptions with 10 different rewards controlled by item.
The upper limit of redemption transactions per second is also 15 for rewards that are controlled by item, as displayed in the following graph:
Updated about 1 year ago