Ivideon cut costs and improved reliability by migrating to Yandex Cloud

Background

Ivideon is an international company that develops cloud solutions for video surveillance and analytics. Its services are used by both businesses and private customers. More than 3 million cameras around the world use the system. Their solutions are based on machine learning, video processing, and cloud computing technologies. Ivideon solutions are easy to integrate with various devices and platforms.

Within a few months, the company moved the main components of its platform from the infrastructure of a foreign cloud provider to Yandex Cloud using managed services. This transition was beneficial: deployment and maintenance costs for new infrastructure stayed in line with the plan, and financial risks were reduced.

Meet data localization requirements and ensure the smooth operation of services

Since 2011, Ivideon has been operating in the cloud, taking easy connectivity, flexible scaling, and high reliability into account. In the beginning, the system was hosted on the servers of one of the popular providers. Over time, however, dependence on a single provider and the threat of service unavailability began to raise serious concerns. Since 85% of the company’s customers are in Russia, and Federal Law No.152 requires local hosting of data, the company decided to move the platform to the Russian cloud while keeping the same level of service quality. At the same time, it was important to maintain data integrity and the service availability.

The Ivideon team studied several cloud platforms and compared them in terms of ease of integration, launching services, and price/quality ratio. The key requirements were quick access to computing resources, convenient tools for launching services and managing infrastructure, as well as compliance with information security regulations, including Federal Law 152-FZ and other requirements.

At the same time, Ivideon was trying to maintain the level of costs and the quality of customer service. As a result, the company chose Yandex Cloud: the platform offered stable services, a convenient console, an SLA of at least 99.95%, and competitive prices. An additional advantage was the localized support service.”

Fault-tolerant Yandex Cloud infrastructure for a high-load video analytics service

The Ivideon team developed the solution on a modern technology stack: a Python^™ and C++ backend using Cython, a React frontend with Backend‑for‑frontend in PHP, and mobile apps for Android and iOS, taking the peculiarities of each platform into account.

Ivideon independently carried out the migration with the support of Yandex Cloud specialists. The transfer took six months and went through three stages: planning, data migration, and optimization. The main task was to maintain uninterrupted service operation, which the team achieved thanks to Yandex Cloud’s detailed planning and automation tools.

Ivideon developed the video analytics program independently, and licensed facial recognition and car number recognition technologies from Tevian. The solution now has a hybrid infrastructure. Tevian is deployed on the former provider’s servers, and where videos for customers outside Russia are also stored. The CI/CD pipeline was built on local versions of GitLab and Jenkins. Customer data is stored in MongoDB and ClickHouse^® databases deployed in an external data center. Key product components have been transferred to Yandex Cloud: API, billing, web app. They also connected Yandex Cloud Postbox for email newsletters. The solution involved Yandex Compute Cloud — a total of 90 virtual machines. Containerized services were located in the Yandex Managed Service for Kubernetes^® cluster. Yandex Object Storage stores videos from Russian customers.

Ivideon is currently operating the solution and continues to optimize it.

Cost optimization and high service availability

Ivideon completed its migration to Yandex Cloud and reduced its infrastructure costs. The level of availability and resilience against failures has increased, and flexible scaling is now possible. The company reduced the time it takes to launch video surveillance for customers.

In the near future, Ivideon plans to expand the use of Yandex Cloud when developing new video analytics products, e.g., automating a MaaS solution based on artificial intelligence and using cloud resources for international pilot projects if their own capacities are insufficient. In addition, the team intends to add new cloud-based features to the platform.