Aiven supercharges streaming giant Vidio’s engineering teams

CASE STUDY

By having Aiven manage our databases, our team can focus on building our core business.

Andy Librian

Technical Expert

Overview

Vidio is Indonesia’s leading video streaming service, with more than 60 million monthly active users consuming 82 livestreams and over a million video-on-demand titles. The company offers a rich variety of content including local and international entertainment and live sports, video-on-demand services, and pay-TV channels. Due to their volume of content selection and watchers, Vidio is one of the top apps downloaded by Indonesians, the world’s 4th largest population.

Uses: Aiven for PostgreSQLAiven for RedisAiven for OpenSearch

Cloud: Google Cloud Platform

Industry: Media & Entertainment

Year founded: 2014

Company size: 400 employees

Products used
PostgreSQL
Redis
OpenSearch
Hosted on
Google Cloud Platform

The Challenge

Vidio needs to guarantee low-latency streaming for end users while customizing the user experience. This is absolutely critical for success in the hyper-competitive streaming market.

Yet with such a large customer base and so many sophisticated services on their network, this is easier said than done. Vidio’s technical team was spending far too much time on database management, up to the point where it was pulling them away from developing services — which is what they were originally hired to do.

“Database management is very important work and key to our overall mission, but it doesn’t directly contribute to the bottom line”, says Andy Librian, Technical Expert at Vidio.

Still, recurring problems with data pipelines and scalability have tangible consequences for streaming companies. In the past, Vidio would sometimes reach maximum capacity during live events like sports matches or music festivals.

“When we had a capacity or scalability problem due to a bottleneck in the database, within seconds we would be hit with customer support tickets and social media criticism”, Librian says.

They needed to solve these problems for good, while at the same time giving their developers a chance to take on higher-level work.

“We wanted to find a managed service to free us from all data plumbing", Librian explains. “The solution also needed to be simple, or it really wouldn’t have had any value.”

The Solution

After reviewing multiple options, the company turned to Aiven.

Vidio is now using Aiven for PostgreSQL, a fully managed SQL database that can be deployed in any cloud. PostgreSQL is the company’s default choice for transactional data — such as user data, transactions, subscriptions and videos.

With Aiven, Vidio can deploy new services at a rapid pace.

“When there is a need to create a database for a new service, we know that we can spin it up in a few minutes with Aiven”, Librian says.

In addition, Aiven also provides Vidio with Redis, which is an open source in-memory data structure store. Vidio uses Redis for caching, ephemeral data, and its Pub/Sub feature for chat.

Vidio case-study solution scheme

The Outcome

“Thanks to Aiven, we can now scale our databases to accommodate rapid fluctuations in user volume”, Librian explains. “This means we’ve been able to improve output and customer experience — while also gaining the ability to roll out new features at a much faster pace.”

Aiven shined during the early stages of the pandemic, especially when Vidio saw huge spikes in traffic (15x over a few hours) and their subscribers grew 3x in April.

“Database management used to be a huge bottleneck for us”, Librian adds. “We could scale our application servers, but we couldn’t scale our databases quite as easily. Now we are able to scale using multiple replicas. It takes only two clicks to add a new replica on Aiven.”

In fact, Vidio is able to practically scale its databases in real time with Aiven.

“We configure our backend application to connect to replicas for read-only requests (http GET requests)”, says Librian. “Then we use a client-side/sidecar proxy to load balance the connections to multiple replicas. When we need more capacity, we can spin up a new replica, attach it to the discovery mechanism, and HAProxy will be able to pick it up in less than 10 seconds. It works seamlessly without restarting the backend services or the proxy.”

Case study vidio outcome

To their satisfaction, the developers haven’t had any issues with capacity and scalability since using this technique with Aiven.

“Now that we are able to scale our databases, we are more confident to run big events and no longer worry about the customers’ complaints”, Librian says.

In addition, Vidio benefits from Aiven’s rapid response support.

“We really enjoy Aiven’s hands-on customer support. It’s a much better experience than what we’ve seen with other big cloud vendors”, Librian says.

For Librian and the rest of the team at Vidio, Aiven is proving to be a key partner.

“We know that we can count on Aiven”, says Librian. “We’ll use them for existing and new databases as we continue to grow.”

We really enjoy Aiven’s hands-on customer support. It’s a much better experience than what we’ve seen with other big cloud vendors.

Andy Librian

Technical Expert

Check out our other case studies

CatalystTalon.OneJobCloudPafAlef EducationNetSpyGlassAdeoOVO EnergyComcastAivenNorautoSpareOmetria

Let‘s connect

Aiven for Apache Kafka, Aiven for Apache Kafka Connect, Aiven for Apache Kafka MirrorMaker 2, Aiven for M3, Aiven for M3 Aggregator, Aiven for Apache Cassandra, Aiven for OpenSearch, Aiven for PostgreSQL, Aiven for MySQL, Aiven for Redis, Aiven for InfluxDB, Aiven for Grafana are trademarks and property of their respective owners. All product and service names used in this website are for identification purposes only and do not imply endorsement.