Resilience

We designed SearchPilot with security and resilience as a top priority from the outset. The system is engineered to be fast, robust, resilient and secure.

Technical documentation, designed to be shared with engineering and operations teams, is available upon request.

You can read more about our approach to security on our security and compliance page.

Approach to resilience

SearchPilot operates in 7 AWS regions, three in the US, two in the EU, then one in both Tokyo and Singapore.

SearchPilot is designed with multiple levels of resilience and failover in place. In each AWS region traffic is load balanced across an auto-scaling group of servers. At times of high load new servers are automatically provisioned to maintain a certain level of redundancy in platform capacity. Each server can fallback to being a transparent proxy in the event of an application error.

By default we configure the system to route around SearchPilot entirely in the unlikely case of catastrophic failure or complete outage of AWS across multiple availability zones and regions. This can happen almost instantly via Amazon’s Route53 service as well as at the CDN level for sustained outages. Under any of these circumstances your site stays up and available but without the SearchPilot enhancements and tests.

Each of our 7 AWS regions have redundancy and can handle millions of requests per minute. Currently, we handle billions of page views per month.