Hornet Networks is the world’s leading queer tech company, providing a digital home for LGBTQ+ people to safely connect with each other — anytime, anywhere. Amplifying the radical, affirmative power of the queer community with cutting-edge technology, Hornet serves over 35 million people worldwide with a wide range of features from video to chat that makes it easy and safe for queer people to share experiences and be authentic in a supportive community.
We recently spoke with Matthew Hirst, Head of Engineering at Hornet Networks, and Nathan Mitchell, Lead DevOps Engineer, about the company’s origin, direction, and how DataStax supported its mission with DataStax Luna initially and now with DataStax Astra DB.
1. Tell us about Hornet Networks and what sets it apart.
Since 2011, Hornet has been committed to empowering queer people and cultivating meaningful LGBTQ+ relationships, both among our users and internally. We are a global company with a remote working culture. Engineering is centered in Cape Town, South Africa, and we also have offices in West Hollywood, Prague, and Hong Kong.
Hornet has disrupted dozens of markets such as France, Russia, Brazil, Turkey, and Taiwan, to become the number one queer app and is rapidly expanding its sizable user base across Europe and the United States. We saw a need in the LGBTQ+ community for a safe online space where individuals could meet to share ideas, experiences, and interests.
Our video livestreaming platform, Hornet Live, just officially surpassed 100 million streaming minutes and has proven to be a reliable income source for queer content creators. We are constantly innovating and evolving our social network to provide more value to our unique user base.
2. What are some of the biggest challenges that your company faces when it comes to data?
We wanted to support the launch of a new social networking application that would complement the company’s existing dating app. The new app, SPACES, provides a safe environment for members of the LGBTQ+ community to meet each other around shared interests and locations.
Accessibility and safety were chief concerns of our users. That had big implications on how we manage data, servers, and the network. We needed to build a welcoming space, available around the clock for real-time interaction. We had to design our back-end infrastructure to support how this new community app would grow over time, as well as provide that real-time experience.
We have users around the world who expect to access SPACES any time—day or night. So, we must build the app and its infrastructure to support that reliable performance and availability. As a lean engineering team, we were also acutely aware that managing the infrastructure steals time from other projects. So, we needed to minimize—or even eliminate—Hornet’s role in day-to-day maintenance.
Finally, we had to think about scale and its implications on data management, infrastructure, and cost. We expect SPACES to grow dramatically in user numbers and volume of interactions. We need to efficiently store and access data, properly size the clusters and other infrastructure, and at the same time keep costs down. Ideally, we want just the right amount of infrastructure we need at any given moment.
We came to realize that all these challenges could be met by DataStax Astra DB.
3. Why did you choose to use Apache Cassandra?
Hornet is a long-time Apache Cassandra user because of its ability to handle large volumes of data very efficiently, with very low latency and read times. Our apps connect people with common interests around the world for social networking and dating. Messaging and feeds are central to the apps and must be real-time to have relevance. Cassandra is a great platform for what we need, and we’ve used it for our primary messaging and feed clusters for several years. Frankly, it is one of the most stable components in our environment.
4. Why did you start working with DataStax?
Our Cassandra environment was humming along for years, running just perfectly. Then one day the feed cluster died. We replaced the cluster and, after a bit of research, asked DataStax to help get that new cluster online.
DataStax checked the health of all our clusters and tuned them up with some configuration changes, helping us get more out of each one. It was incredibly valuable, providing us the option to extend the life of the environment by three years. The work was covered by a subscription to DataStax Luna, which gave us three days of consulting plus a year of support.
It was a welcome service for our internal IT operations team, which had been swamped managing the Cassandra clusters and trying to keep them up to date. Each update was a major project that took time and resources and had to be performed while also maintaining essential, real-time functionality. By subscribing to Luna, we were able to free our team to focus on other aspects of our data infrastructure.
5. What attracted you to DataStax Astra DB?
When we were building our SPACES app, we reconsidered our infrastructure approach. We had positive experiences with DataStax and were intrigued by Astra DB. As a cloud-native database-as-a-service built on Cassandra, it had many potential benefits for Hornet—most notably automatic scaling and pay-as-you-go pricing. Hornet could effectively scale costs upon adding users, rather than investing heavily up front.
We started with one cluster to support development, allowing us to get familiar with Astra DB. Following production deployment of SPACES, we turned our attention to migrating our existing clusters that support the dating app.
With the use of a proxy server, all new application data was written to both the dating app clusters and to the new clusters on Astra DB. Concurrently, historical messaging and feed data was migrated to bring the new cluster up to date.
Once the migration, using the DataStax Zero Downtime Migration Tool, was complete, our engineering team was freed from all maintenance activities related to Cassandra. The biggest win for us was that nothing changed; response time is the same and there is no difference in user experience, which is amazing.
Of particular value are the automatic scaling and pay-as-you-go features of Astra DB. As SPACES grows in popularity, Hornet will simply pay for additional usage as needed rather than invest in additional hardware in anticipation. We don’t need to worry about disk space and can throw more data at it because of the better data compaction DataStax achieves.
We also looked at our costs for managing our internal clusters compared to moving to Astra DB and estimate that we will see savings of between 10 and 20 percent on our costs. We now have immediate support in the unlikely event of downtime, and do not need to have a specialist DBA on staff to keep this critical database tuned.
6. What advice would you have for other organizations considering DataStax?
We started out small and quickly gained confidence to go all in. We worked closely with DataStax, which gave us confidence in their team. And now we can rest assured that our Cassandra-based solution will be up-to-date, properly sized, and tuned. Given our great experience, we recommend a similar approach for others to follow.
Read the full case study to learn more about Hornet and why the company chose DataStax Astra DB to support the expansion of its queer social network services.
