On one of the most critical shopping days of the year, CJ’s disaster recovery plan kicked in due to Amazon Web Services (AWS) and Google Cloud outages. Our pre-planning ensured that our platform and client programs continued to run without any interruption while other companies across the web experienced downtime.
As a marketing technology company, how do you prepare for unpredictable surges in volume during a peak shopping period—as we’ve seen during this global pandemic?
The shift to ecommerce has made many companies reflect on their preparedness to support the increased demands on their technology and service providers. While the topic of technical architecture may not be as exciting (to most) as a new product launch, a solid disaster recovery plan is critical—especially during times with planned and unplanned peak volume. Having a plan in place ensures that products continue to perform flawlessly with uninterrupted service.
Why CJ's System Held Up
CJ’s tracking infrastructure is designed to process high volumes of data with minimal delay, with the ability to immediately scale when traffic needs increase dramatically. Our high fault tolerance and intentional redundancies allow our systems to continue operating properly in the event of a failure in order to prepare for planned and unplanned outages. As part of our well-established and agile Disaster Recovery Plan, CJ’s system has a staggering 99.99% uptime for tracking—even during major industry downtime, such as the recent AWS and Google Cloud outages. Part of this preparation includes being ready for unprecedented scale or points of failure through a combination of on-site applications served from multiple locations that are geographically separated. This strategy is paired with DNS-based routing and load balancing, which provides near-immediate load times for call network traffic and ad delivery across the globe.
Our cloud-based micro-services strategy also allows each part of our platform to be architected independently with appropriate fail-safes, further mitigating the risk of full-system outages or delays—unlike monolithic architecture of the past, where if one part failed—everything failed. CJ’s independent micro-services are all deployed to multiple availability-zones as well as multiple regions as active-active architecture, which means all components reside on the same system, and additional systems are configured similarly and provide similar functionality for failover.
Safe and Sound—As It Should Be
This holiday shopping season put many companies’ fail safes and disaster recovery plans to the test with failures across multiple cloud bases and other service providers. While CJ leverages cloud-based services, our architecture has built in fail-safes and redundancies that allowed us to continue providing uninterrupted service during peak holiday shopping.
Q4 is already stressful enough—we’re so grateful to have a stellar team with such an effective plan. This is how CJ engineering saved the holidays!