• Hyperscale cloud providers report a surge in demand for both their on-demand infrastructure services and productivity and collaboration services
• Past experience is so far helping guide capacity planning but there are still some limitations that are impacting performance as some users and applications involved in critical healthcare and emergency services get priority
The COVID-19 pandemic is driving businesses in virtually every field to deploy a remote workforce model overnight. This shift comes with immediate need for cloud-based productivity and collaboration applications. Organizations are also looking to virtualize other elements of their infrastructures, and thus requiring more cloud capacity to support these changes.
Microsoft disclosed in a blog updated this week that in one month it experienced a 775 percent increase in Teams’ calling and meeting users in Italy. The company noted that globally Microsoft has seen a big jump in Teams use from the now more than 44 million daily users who in one week held over 900 million meetings. Microsoft Windows Virtual Desktop use has tripled, and use of its Skype conferencing and calling platform shot up 70 percent in March.
The upward tick in cloud usage is raising concerns that on-demand platforms might buckle. Microsoft has said that it is granting preferential capacity and access to certain users including first responders; emergency routing and reporting applications; medical supply management and delivery systems; emergency alert applications used by first responders, healthbots, screening applications and websites; and health management applications and records systems.
To maintain sufficient performance/resources, Microsoft has imposed temporary capacity caps and other limitations for other customers. Some users in both the U.S. and Europe noted they had hit capacity limits and weren’t able to spin up new virtual machines.
Though Microsoft said resources may be constrained at times, the company reports no major outages. Microsoft is also accelerating the deployment of new capacity to support clients in regions that are in acute need of more cloud resources.
The other hyperscalers are also taking measures to prioritize critical user groups and ensure stable operations. Amazon Web Services’ (AWS) disaster recovery plan includes pandemic response policies and procedures specific to issues associated with infectious disease epidemics. The provider incorporates everything from alternative staffing models to shift critical processes to a different region to the activation of a crisis management plan. AWS applies learnings from previous spikes in demand to ramp up capacity as needed.
Google Cloud deployed remote customer service agents and added support for peak periods. The provider also emphasized that it has “rigorously tested the resilience” of its infrastructure and processes to ensure optimal and stable performance.
While it is too early to tell how ongoing shifts in cloud capacity demand will impact service availability and performance, one thing that is clear is how critical this infrastructure is to business and government. It will also be interesting to see if some of the currently temporary remote operating models might become permanent.