Exceeding the highest standards
Our presence in 55+ data centers across 15 regions is unmatched. Wherever you are in the world, we’ll make your search ultrafast and reliable!
DSN - Distributed Search Network
Easily distribute your search engine on up to 15 regions worldwide.
Our Distributed Search Network allows you to automatically replicate and synchronize your data across 15 regions, and search queries are automatically directed to the closest data center.With our DSN, network latency is dramatically reduced.
Unique high availability
We are consistently striving for 100% uptime and take our redundancies quite seriously.
Your data is replicated on at least 3 high-end servers that can independently handle each query.Even if 2 servers are down, your search engine is up and running.
These 3 servers can be hosted by up to 3 different providers, across 3 data centers, through 3 autonomous systems, enabling multi-region reliability in a single region.Even if an entire data center is down, our provider redundancy ensures service availability.
We have introduced a DNS fallback strategy that relies on a second completely independent infrastructure.A broken DNS resolution will not stop your queries from reaching our API!
Ensuring business continuity
Information security is at the top of our minds since day one.
- Everything is considered as unsecure, even communications within the same rack are encrypted.
- Only HTTP and HTTPS ports are publicly opened.
- Servers are always running on the latest stable version of nginx web server.
- Infrastructure is protected by multilayer access control.
- Servers located in datacenters built following the industry security standards.
- A continuous security testing through a public bounty program is opened on HackerOne.
An incremental backup of your data is built every couple of hours.
For a faster service recovery, a client-side encrypted copy of the data is also sent to the closest AWS S3 region.
We are passionate about sharing how we build and fix stuff.
Our team wrote about the journey to build our insfrastructure.
When a major issue occurs, we publish a post mortem on our blog to publicly open a discussion and share our insights.
Our API monitoring status page provides detailled information about performance and outages from 50 locations worldwide.