Network outages intensified and affected more users and services in 2022

A year in outages: 2022 saw various high-profile outages take down apps and services. Many of these outages global in scale, indicating continued stress on an overburdened infrastructure that’s become unmanageable.

Outages were caused by human error, bad software updates, and carelessness on the part of system administrators. Let’s look at the litany of outages that took down services in 2022.

January

• Apple started the year with a monumental iCloud outage that took down Apple Web Apps, iCloud Backup, iCloud Mail, iMessage, and iCloud Photos, which were inaccessible or offline across the world for two days.

April

• Thousands of YouTube viewers and PlayStation Network gamers were booted offline as an overloaded network infrastructure continued to take its toll on high-bandwidth services. 

May

• Xbox Live spoiled gamers’ weekends as a massive outage prevented players from launching cloud games, making purchases, and signing into accounts.

June

• An outage hit 19 of Cloudflare’s data centers, resulting in Amazon, Twitch, AWS, Steam, Coinbase, Telegram, Discord, DoorDash, GitLab, and others going dark for some users. Cloudflare’s outage affected roughly 50% of all global HTTP requests handled by the content delivery network.

July

• A heat wave knocked out Google and Oracle data centers in the UK, exposing the fragility of thermally challenged data centers during summer months and extreme weather spikes.
• In Canada, a nationwide Rogers outage took out a quarter of broadband, payment systems, transport, government access, healthcare, and even 911 calls. 

October

• South Korean super app Kakao had a nationwide outage that disrupted communication, prevented mobile payments from going through, and stranded passengers trying to use ride-sharing services.
• WhatsApp, Meta’s messaging app, suffered a significant outage for two hours, which resulted in spiking global searches for alternative messaging apps.
• Meta’s services continued to have outage issues, with Instagram booting off thousands of users and suspending their accounts for no apparent reason. 

December

• Rackspace had a multiday outage that took down email and Microsoft Exchange services for thousands of users, exposing the danger of relying on third-party hosting companies.

Key takeaways: Recent outages aren't just more frequent—they have also been taking longer to resolve than previously, indicating that massive growth is quickly becoming unmanageable even for companies with considerable resources.

Consumer broadband data consumption has reached a milestone of more than half a terabyte each month, an increase of 165% compared with the 203 GB monthly consumption in 2017.

• We’re seeing Big Tech lumping their various services, apps, and systems into a monolithic infrastructure that takes everything down when it goes offline. 
• Rogers’ patchwork infrastructure cost the company C$10 billion ($7.7 billion) to repair its network and physically separate wireless and internet services that have been layered over each other for decades.
• The US’ existing national long-haul fiber infrastructure was built to support systems, broadband demands, and networks that existed 20 to 30 years ago. Simply adding servers and towers to handle more traffic and users seems to be causing even more complex problems to the infrastructure. 

What can we expect in 2023? Remote work will continue to push consumer broadband services to their limit, especially as streaming services, IoT devices, and online gaming, as well as remote work and education increase exponentially.

• Affected businesses could shift away from cloud computing and hosting services, or at least revert to hybrid and on-premises solutions to better diversify and maintain greater control and security of their services. 
• This solution would dampen growth in the cloud and infrastructure markets but could also push providers to undertake generational system and infrastructure upgrades.