Public Cloud Down Again? Predictions for 2021.

This post is also available in: Japanese

I thought we had heard the end of availability issues last year (see my blog from last year). After all, reliability and having enough processing horsepower to keep applications available across failures is one of the big benefits of migrating to the cloud or developing applications natively in the cloud.

But NO. This year, the Amazon AWS outage outdid everything we have seen before; many services — such as Roku, Vonage, Adobe, Washington Post, Flickr, Autodesk, to name a few — were impacted for 8+ hours. Such service disruption often leads to poor customer experience, and attackers and hackers know this and use a broad set of techniques to cause harm. Will the move to the cloud slow down with this outage? In short, NO. But organizations will be more willing to hedge against having all computing eggs in one vendor basket.

With that said, here are my predictions for 2021:

Continued Push to the Cloud

• Organizations will evaluate multi-vendor and multi-cloud options to hedge against outages at their primary cloud providers;
• Revenue-impacting business applications will be hosted across multiple providers, and companies will require service level agreements and demand high availability solutions;
• Easy availability of sophisticated hacking tools and bots will force many to pay for keeping service denial attacks out of corporate / virtual private networks;
• There will be more willingness to invest in education to address issues such as phishing and social engineering that play a large part in human failures.

More Remote Workers

• Further investments in solutions to enhance end-user experience by incorporating caching, compression, WAN and front end optimizations;
• Move toward a zero trust environment to ensure that the applications are accessed by the right users who are authorized and authentic;
• Further deployment of multi-factor authentication, single-sign on, client authentication, removal of unsafe ciphers and moving to TLS 1.3;

[You may also like: The Future Of Work: The Hybrid Workforce]

Scraping and Bot Attacks on Applications Will Continue to Increase

• Organizations will evaluate and invest in better security mitigation technologies, including bot protection, API security, application security, and data leak prevention technologies;
• Further investment in visibility and forensic tools in cloud to gaining actionable visibility for management, monitoring, auditing, compliance, forensics and troubleshooting.

A Continued Lack of Multi-Cloud Networking and Security Expertise

• Lack of necessary human expertise will force more configuration automation to deploy corporate networking and security policies – now across multiple cloud environments;
• Improved automation and orchestration tools will emerge to help deploy application scalability, monitoring, security, optimization constructs across multiple clouds;
• Many MSPs will offer multi-cloud expertise to customers willing to pay someone else with the expertise.

[You may also like: Here’s Why We Need to Stay Safe in the New Contactless Economy]

The Cost of Cloud Deployments Will Again Become a Concern

• Now that the metered and pay-as-you pricing model has delivered some pricing shocks for organizations that were forced to move to cloud-only deployment due to the pandemic, organizations will look to reduce cost of compute and licensing;
• Elastic and flexible BYOL models across multi-cloud environments will gain favor with MSPs and large organizations;
• Costs of operating in the cloud will further increase as customers pay for both security and visibility in addition to compute of workloads.

The COVID-19 pandemic has accelerated the transition timeline to the cloud for many organizations. However, the outages at many of the largest cloud service providers and recent hacking and ransom attacks highlight availability, scalability and security challenges that must be addressed to keep both customer data and businesses safe and available.

Note: A version of this article first appeared on VMBlog.com.