Global cloud research reveals variations in major providers’ performance

Features

After Microsoft Azure’s and Google Cloud Platform’s recent outages, this second annual report on the global cloud market highlights some big differences between the big five.

In its second Cloud Performance Benchmark, ThousandEyes compared global network performance and connectivity differences between the five major public cloud providers – Alibaba Cloud, Amazon Web Service, Google Cloud Platform, IBM Cloud and Microsoft Azure. ThousandEyes is an enterprise software platform that measures how Internet, cloud and other third-party dependencies impact end user digital experiences.

Reliance on public internet

If found that some cloud providers rely heavily on the public Internet to transport traffic instead of their backbones, which can affect the predictability of their performance. Google Cloud and Azure rely heavily on their private backbone networks to transport their customers’ traffic, protecting them from the vagaries of the public Internet.

AWS and Alibaba Cloud rely heavily on it for most of their transport. ThousandEyes reckons this results in greater operational risk as performance is less predictable. IBM takes a hybrid approach that varies regionally.

Regional variations

Latin America and Asia have the highest performance variations across all clouds, whereas in North America, cloud performance is generally comparable. Hence decision-makers should check the detailed findings to choose the best cloud provider on a per-region basis.

AWS Global Accelerator

AWS Global Accelerator was introduced in November 2018 whereby customers pay a fee to use the provider’s private backbone – rather than the public Internet, which is AWS’ default practice. The Accelerator does provide better performance in many places globally, but there are examples where the Internet performs faster and more reliably than the Global Accelerator, and in others there is nothing to choose between the two.

ISP choice affects cloud performance

Businesses looking to gain possible every performance edge should consider which ISP they choose for broadband and base that selection on which cloud provider they depend on most heavily. There are performance gains and losses depending on which broadband provider businesses use to connect to each cloud.

The Great Firewall of China

Despite Alibaba’s origins in China, it experiences packet loss when crossing through China’s Great Firewall just as all the other cloud providers do: it does not enjoy any preferential treatment. The report suggests enterprises serving customers in the region might consider Hong Kong as the base to serve them from. Alibaba Cloud traffic experienced the least packet loss from Hong Kong to China, followed by Azure and IBM.

Archana Kesavan, research author and director of product marketing at ThousandEyes said:

"When businesses need to decide which cloud provider best meets their needs, one metric that's notably missing from their assessments has been performance data, mainly because it’s never been available or has, at best, been myopic. The second annual Cloud Performance Benchmark gives businesses that comparative data. Understanding cloud performance is essential for planning and for ongoing measurement so you can be assured that you're providing customers and employees with the best possible performance."

No cloud is created equal

“Historically, most IT professionals select cloud providers based on factors such as price or proximity to users, but often overlook the underlying network architecture of the cloud providers, which can have a significant impact on performance.

"The 2019 ThousandEyes Cloud Performance Benchmark shows just how widely performance can vary from the top cloud service providers and reinforces the point that businesses should remain diligent about collecting network intelligence and measure their own performance to ensure they are getting what they expect from their cloud providers,” commented Zeus Kerravala, Principal Analyst, ZK Research.

The inaugural Cloud Performance Benchmark, published in November of 2018, benchmarked the three biggest cloud providers (by market share) AWS, Azure and Google Cloud Platform. Alibaba Cloud and IBM Cloud are new this year, meaning this year’s report has no prior comparison points. Nevertheless, it turns out there are notable differences between each cloud in terms of overall performance and connectivity architectures.

Key findings, by cloud provider include:

•    AWS generally demonstrates low latency and made some minor improvements over the last year. Its performance predictability metrics, however, noticeably improved, with its most dramatic change in Asia, which showed a 42% reduction in variability over last year. It still has lower performance predictability than when compared to Azure and Google Cloud Platform, though, due to its heavy reliance on the internet rather than its own backbone.

•    Microsoft Azure continues strong network performance through using its own backbone to carry users’ traffic to cloud hosting regions. Big changes from 2018 include a 50% improvement on average in performance predictability in Sydney, while in India, there was a 31% decrease in performance predictability. Despite a slight decrease year on year, Azure lead in performance predictability in Asia when compared to the other cloud providers.

•    Google Cloud Platform continues to favour its own backbone, but has some significant gaps that haven’t been resolved since last year’s report. Traffic from Europe and Africa takes 2.5-3 times longer to get to India, as it is routed around the rest of the world instead of taking a direct route. The company also decreased visibility into its internal network, making it harder for its users to understand network paths and performance.   

•    Alibaba Cloud has comparable performance to other cloud providers and closely resembles AWS in terms of connectivity patterns, region locations, and even region naming constructs. Like AWS, Alibaba Cloud relies heavily on the internet instead of its own private backbone for transporting most of its users’ traffic. Uniquely, traffic between Alibaba’s clouds regions is not contained within its own cloud infrastructure but exits Alibaba Cloud, traverses the internet and then re-enters Alibaba Cloud.

•    IBM Cloud performance is comparable to the major players. The company takes a hybrid approach to traffic delivery, fluctuating between using its own private backbone and the public Internet, depending on which regions users’ traffic needs to access.