The Cyoda platform is dynamically horizontally scalable, truly distributed, with no single point of failure in all aspects, processing, storage and reports/queries. There are no limits to the quantity of data it can store, the transactions it can run simultaneously, the processing through the rules engine, the queries are broken down and distributed and an inbound data feed is distributed across the platform. It is this true distribution of all aspects and its tolerance to fault/failures that allows Cyoda to perform virtually all data processing and reporting requirements for the world’s largest data-intensive organisations such as a tier 1 investment bank, quickly accurately and reliably.

The term scalable is bandied about very often in technology  but  there are different types of scalability:

Cyoda is horizontally scalable and elastically scalable: Scaling a system by adding more commodity nodes. These nodes can be physical servers (bare metal blades) or virtual servers, rented for just an hour when needed. Cyoda can scale up or down dynamically whilst running processing and reports, there is no need to have downtime.

vertically scalable: Moving a system to a larger machine, for example, a relational database. This involves shutting down the system and moving the system to the new machine with application downtime. Ultimately there is a limit to how much a system can scale when it  is bound to a single machine.

Horizontally scalable systems are typically distributed systems, but some are Slave Master and mirrored Slave Master and a few are truly distributed with no single point of failure, no bottlenecks. Dynamically scalable is the ability to scale up and down the system whilst it is running.

Many systems have aspects of scalability, a NoSQL data store may be scalable in terms of its ability to store data, scale in its ability to handle a large number of small queries but may not scale for a single large query or a system may be able to scale up its compute power through utilisation of a compute grid but still be dependant on a single relational database.

The Cyoda platform can be run on in-house commodity servers or on a cloud provider such as AWS. Cyoda’s ability to be dynamically scaled coupled with AWS’s pay for the time used means you can scale the system up and down in seconds to meet fluctuating demands.

More Info

Further reading:

Overview

STP Processing

Easy Integration

Distributed Queries

Data Consistency

Technical Overview