The Need for Operational Synchronization
Filed under: Business Intelligence, Data Analysis, Data Integration, Metadata, Recommendations, Replication
Over time, organizations have employed a variety of strategies for managing data assets in accordance with the specific needs of the different business applications in operation. For the most part, applications were designed to achieve specific objectives within each business function. Correspondingly, any data necessary for the business function would be managed locally, while any data deemed critical to the organization would be subsumed into a centralized repository.
Yet this approach to centralizing data has come under scrutiny. Organizational systems have become more distributed, relying less on a single centralized system and more on collections of interacting systems. A growing number of systems are engineered to explicitly or implicitly share data. In turn, there is a need for high availability and consistency of data sourced across geographically diverse locations. And when each of these individual systems only contains a subset of the total collection of data about relevant data entities, bridges must be developed to align operational and transaction data to support operational processing as well as downstream reporting and analysis.
In other words, modern enterprises must incorporate methods for what is called “operational synchronization” – a practice of employing tools and technology to maintain acceptable levels of currency, consistency, and completeness of a composite view of information blended from applications across a set of distributed and heterogeneous applications and data sets. For example, an organization might have a variety of customer-facing business processes, each supported by its own application and corresponding data store holding some aspects of customer information. Enabling a customer data visibility demands operational synchronization of the “partial views” into one comprehensive view.
Operational synchronization poses a challenge in balancing data availability and consistency with maintaining acceptable levels of system performance. Therefore, some key characteristics of operational synchronization include:
- Real-time, rapid data integration – Modifications or changes to operating systems must be captured rapidly and quickly communicated to the other cooperating data environments and loaded to ensure observance of defined levels of service.
- Non-invasiveness – the methods for operational synchronization must not interfere with business application execution, but rather detected and communicated in the background.
- Heterogeneity – The coordinated systems are likely to have been designed and implemented at different times by different teams on different platforms, so the operational synchronization must support heterogeneous sources as well as targets.
- Bi- or Multi-directionality – Even though many environment will only want to push data out of one system, there are many scenarios in which data flows among a cohort of coordinated business applications to provide the comprehensive view.
- Transaction Integrity & Data Consistency– Proper synchronization is improved when the technical solution already has built-in capabilities for embedding rules for integrity validation and transformations, allowing you to ensure the integrity of transactions.
In the upcoming posts we will look at some use cases for operational synchronization and the technical requirements for enabling synchronization.