APIContext has unveiled its 2024 Cloud Service Provider API Report, offering an in-depth examination of the performance of leading cloud service providers in the essential API domain. This annual global analysis scrutinised data stemming from 650 million API calls across over 10,000 distinct API endpoints. These interactions were sourced from a netwrok of over 100 geopraphically dispersed cloud data centers, encompassing major public cloud service providers such as AWS, Azure, Google, and IBM.
The results show that, in 2023, globally only 7% of API Services achieved a 99.99% service availability score, a significant drop from 18% in 2022.
Achieving 99.99% is the benchmark all API service providers should be aiming for, while 99.999% is the traditional telecommunications standard for service availability – equating to five minutes of downtime in a year.
Reasons for the drop in service availability can be attributed to the scale in which the cloud ecosystem has grown by 20% in the past 12 months, making it difficult for cloud service providers to continue scaling hardware at the same rate as consumption growth.
Today, cloud computing comprises most of all internet traffic. With remote working demands as well as the rise of artificial intelligence and machine learning computing, cloud providers are clearly struggling to maintain speed and quality in all circumstances.
Despite poorer overall service availability, other quality API metrics improved, and on balance, the API services observed in this report had a positive overall performance, as evidenced by Cloud API Service Consistency (CASC) scores reaching above 8.00.
CASC scoring is patented by APIContext and indicates the quality and health of a specific API with a score of 10.00 resulting in a perfect performance.
In fact, 2023 is the best year on record for API performance overall, with 68% of providers achieving a CASC score between 8.00 and 8.99, indicating ‘very good’ performance, while 32% of providers had a CASC score between 9.00 and 9.99, attaining an ‘excellent’ quality rating.
Mayur Upadhyaya, CEO of APIContext: “This year’s Global Cloud API Performance Report underscores a critical gap in the current landscape of monitoring tools.
“The observed degradation in service availability reveals that the majority of cloud service customers are not adequately monitoring the quality of service, with a predominant focus on cost rather than performance.
“Traditional synthetic monitoring tools are simply not designed for the complex authenticated API workflows that drive the majority of today’s online transactions.”
Other key report highlights:
- API Cloud performance varies by geography: In terms of API speed, North America is the fastest region in 2023, 500 milliseconds (ms) faster than Oceania which ranked last in fifth while Europe placed second.
This is expected because Europe and North America have robust, reliable telecommunications networks and higher performance expectations than in regions with slower speeds.
- Four of the top five locations for Time to Connect are Amazon AWS locations: This is the duration it takes an API to process a request and return a response to the client.
Amazon AWS continues to be the flag bearer for optimising network operations and managing traffic flow across systems to ensure speed, which resulted in AWS in Asia, Europe, North America and Oceania all placing in the top 5.
- IBM is the fastest cloud by Total Time: This is the overall time from when the request enters API Management (APIM) until the response is sent back to the client. IBM averages 443 ms while AWS is second with 450 ms. Azure landed at the bottom with 529 ms.
To download a copy of the APIContext 2024 Cloud Service Provider API Report, click here.