Zomato, Swiggy outage: AWS affirms power failure behind services crashing

Two of the top food delivery apps in India, Zomato and Swiggy, went down for about an hour today due to an issue with servers hosted by Amazon Web Services (AWS) in the country. In a statement, the cloud hosting services provider said that a power fault with its virtual servers and data storage media in the Asia-Pacific South-east region led to apps and services hosted on Amazon’s cloud witness downtime for a "brief spell" earlier today. 

AWS reported an issue in their cloud servers from approximately 1:55PM and continuing for about an hour. While specific applications weren't reported, numerous other websites in the Indian subcontinent that are hosted on AWS we're also reported to have experienced about two hours of downtime earlier today.

“Starting at 1:23AM PDT, some EC2 instances experienced a loss of power and some EBS volumes experienced degraded performance within a single Availability Zone in the AP-Southeast-1 Region. Power was quickly restored to the affected instances and EBS volumes and by 1:40AM PDT, the majority of EC2 instances and EBS volumes had fully recovered. By 2:05AM PDT, the vast majority of affected EC2 instances and EBS volumes had fully recovered," an official statement on AWS’ system health tracker page said.

An Amazon EC2 instance, or Elastic Cloud Compute module, is a virtual server that helps apps and websites hosted on Amazon’s cloud platform work. EBS volumes for Amazon refers to Elastic Block Store, which is the storage media for services’ data to be stored on in Amazon’s cloud infrastructure.

In simpler terms, a power supply issue in Amazon’s cloud services in south-east Asia led to apps and websites hosted on AWS to falter and crash.

The instance highlights how issues with cloud platforms can bring popular apps and services down, even as most online services today use multiple cloud service providers for various functions. In December 2021, AWS witnessed multiple major outages that were caused for a wide range of reasons. Other cloud providers are not immune, either – in December as well, a Google Cloud outage brought down numerous popular apps and services such as music streaming service Spotify, social media platform Snapchat and augmented reality game, Pokemon Go.

However, such outages are not a very regular affair, as a result of which various companies do not invest in multiple cloud service platforms. The act of doing the latter is called redundancy, which allows services to shift from one platform to another – in case the other experiences downtime.

Explaining this, Doug Madory, director of internet analysis at American IT services provider Kentik, said, “Usually, redundancy is achieved by using multiple availability zones within a cloud region, or multiple regions within a cloud. It is quite rare that AWS, as a whole, experiences downtime – rare enough that the cost of operating in two clouds far exceeds the benefit.”

Swiggy and Zomato use AWS infrastructure, but it’s not clear if either also use alternate cloud platforms as backup in cases of such downtimes. A spokesperson for Zomato could not be reached at the time of publication of the article. Swiggy declined to comment.