1. Introduction

This diagram shows a generalised view of API performance, using the ""get /accounts"" operation as an example.

Performance is measured using the Time to Last Byte (TTLB) metric. The timer starts when the request is initiated, and stops when the last byte of data is received by the client.

The SLA is defined in the CBUAE Operating Guidelines CBUAE Operating Guidelines at version v1 and is set at 500ms.

The value of 500ms is split 50:50 between the API Hub and a given LFI, as the SLA is shared between both.

Please note the following when reading the diagram:

• The operation for a TPP to get a new Access Token, when required, is out-of-scope of the SLA.

• The TTLB for each API call (e.g. get /accounts) should be 500ms from the time the TPP makes the call to the API Hub to the time the TPP receives the response back from the API Hub.

• The TTLB for each API call made by the API Hub to a given LFI is 250ms to service a call.

• The Bulk/Batch Payments operations to upload a payment file and retrieve a payment report is out-of-scope of the API performance SLA and there this diagram. This is due to the inherent latency of these operations, due to the size of the data transfer being undertaken.

• Response datasets greater than 100 records will be paged. Each page is subject to the API Performance SLA.

2. Sequence Diagram