Building a purpose-built data layer for Web3
Blockchain indexing differs from typical cloud workloads. Instead of handling short bursts of activity, Ormi’s systems run continuously which creates sustained pressure across compute, storage, and networking.
Before adopting UpCloud, Ormi encountered two key challenges.
The first was performance variability. Keeping blockchain data in sync requires consistent performance. In previous environments, Ormi experienced inconsistent CPU performance under sustained load, unpredictable disk input and output during heavy write operations, and latency spikes during traffic bursts. These issues affected the platform’s ability to remain synchronized and reliably serve queries.
“Blockchain indexing requires consistent performance across CPU, storage, and network. Any delay means the system has to catch up before it can serve users.”
Leon Dolinar, VP of Engineering at Ormi
The second challenge was cost structure. Blockchain data grows continuously, and storage is a core requirement rather than optional. Ormi faced storage and data access costs that scaled unpredictably. This required constant effort to optimize infrastructure usage and led to tradeoffs between retaining data and controlling spend.
Over time, these constraints began to influence how systems were designed.
Why UpCloud
Ormi needed a cloud provider capable of supporting sustained, input and output heavy workloads without introducing variability or cost-driven constraints.
Providing better alignment between infrastructure and workload requirements, UpCloud guaranteed consistent CPU and storage performance along with predictable pricing for storage and data access. This allowed Ormi to focus on building reliable systems instead of adapting to infrastructure limitations.
“We needed infrastructure that could handle continuous, data-intensive workloads without forcing tradeoffs in system design.”
Victor Fei, CTO at Ormi
Scalable infrastructure for real-time indexing
Ormi uses UpCloud compute and storage to power its indexing and query infrastructure. The system is designed to handle continuous ingestion alongside high query demand.
The architecture separates indexing, synchronization, and query serving. Indexing nodes ingest and process blockchain data as it is produced. State is updated continuously as new blocks arrive. Query nodes then serve structured data to applications that depend on real-time access.
This setup runs across multiple regions and uses load balancing and failover to maintain availability and performance. It allows Ormi to handle ongoing data ingestion while serving queries under sustained load.
To support this architecture, Ormi relies on consistent CPU performance for continuous processing, high disk throughput for read and write intensive workloads, and stable networking across regions. The platform scales by adding nodes when needed and by increasing resources per node depending on workload requirements.
Performance stability and design freedom
After moving to UpCloud, Ormi saw a clear improvement in infrastructure consistency and system design flexibility.
Disk input and output performance became more stable, and query latency showed less variation. This level of consistency is essential for maintaining synchronization with blockchain networks and delivering reliable data to customers.
“We see higher and more stable disk performance and lower variability in query latency.”
Leon Dolinar, VP of Engineering at Ormi
A more significant change was the removal of infrastructure-driven constraints. With predictable storage and data access costs, Ormi no longer needs to design systems around cost limitations.
This allows the team to retain more data, simplify system architecture, and focus on performance.
“We no longer need to optimize around storage or data access costs, which allows us to focus on building a better system.”’
Victor Fei, CTO at Ormi
This enables Ormi to meet the requirements of customers building production-grade Web3 applications that depend on accurate and up-to-date data.
A collaborative partnership
Ormi has worked closely with UpCloud to ensure that the infrastructure supports the unique characteristics of blockchain data systems. These workloads require consistent performance across compute, storage, and networking layers.
By sharing insights into how indexing pipelines behave under real conditions, Ormi has helped align infrastructure more closely with the needs of real-time data systems.
“We’ve worked closely with UpCloud to align infrastructure with the realities of real-time blockchain data workloads.”
Victor Fei, CTO at Ormi
This collaboration has improved support for high-throughput, data-intensive use cases.
Looking ahead
As Ormi continues to scale, the team is evaluating further expansion with UpCloud. This includes additional regional deployments, performance optimization for high-throughput blockchain networks, and more specialized configurations for large-scale workloads.
