What is CYODA?

Accurate

100% consistency delivered on a robust & reliable architecture,with full data traceability

100% consistent

ACID consistency. Real-time handling of high volume, complex data and business rules, with absolute precision

ACID’ consistency is the holy grail of an operational database. It guarantees the validity of the data in the event of errors or machine failures. No data is available to read unless all operations related to that data are successfully completed.

This is typically important in complex, operational environments, where the accuracy of each individual piece of data within the system is critical. For example, in banking, you would not want accept a payment if the balance was not updated or show funds available from one account when it failed to debit it from another; it’s essential that errors are resolved and the entire transaction be uploaded in full, else the bank’s risk exposures would be incorrect.

Until recently, this consistency has been the preserve of relational databases, achieved by making all updates pass sequentially through a single machine, thereby incurring a substantial cost in terms of performance.

In contrast, ‘big data’ databases are super-fast and scalable, but lack the rigorous accuracy of an ACID relational database. These are typically used for applications which derive value from data in aggregate and the accuracy of each individual piece of data is less critical – e.g. predictive analytics, data mining, social media etc.

Cyoda combines these 2 attributes: ACID consistency across a high performance, 100% distributed architecture.

Always online

No single point of failure; hardware outages will not bring your system down, cause you to lose data or cause errors in system outputs

The traditional method for ensuring ACID consistency typically relies on passing all data through a single machine.  Thus, if that machine fails, so does the entire system.  Cyoda’s distributed architecture means that it is not vulnerable to partial hardware or network failures.  Each part of the system is replicated elsewhere, so in the event of a machine failure, the load will simply be re-distributed, with no impact on the accuracy of reporting outputs.  

Full data traceability

View the full history of any data in the system: from primary inputs, through all transformations, to calculated outputs

Data within Cyoda is immutable which means it is never overwritten.  All updates and transformations are simply stored sequentially alongside the original data. This means that it’s extremely straightforward to derive the history of any piece or collection of data in the system, back to source(s). This includes both the ability to view any updates or changes to primary data, and also to  view how computed data is derived.