Organisations today increasingly are choosing to store their data and perform their analyses in the cloud.
A recent study from Teradata found almost half of organisations are already using cloud-based analytics systems, and data is being migrated to the cloud with 86 per cent, either storing the majority of their data relying in the cloud today or expecting to in the future.
However, cloud is not a wholesale replacement for on-premises data and analytics infrastructure, with nearly two-thirds (64%) of data movement within organisations still occurring between on-premises systems, over one-quarter (27%) moving from on-premises to cloud or vice versa, and only 15 percent happening between cloud-based systems.
With relatively few organisations basing their entire data infrastructure in the cloud, it’s important that organisations consider what role the cloud should play as part of their data and analytics architecture.
Beware of Data and Analytic Silos
Cloud-based data architectures can make it easier to access software, but at the same time can exacerbate the problem of data and analytic silos.
As more line-of-business functions independently purchase their own software, integration between systems becomes difficult, resulting in architecture with data and analytic silos. These silos prevent organisations from deriving the full value of their data and also require costly tools and resources to fix.
Avoiding and breaking down data and analytics silos will thus deliver a number of benefits for the organisation – establishing a “single version of the truth”, reducing storage costs, data transfer times and enabling cross-functional analysis that can generate significant value for the organisation.
Avoiding Cloud-Based Data and Analytic Silos
Single-vendor deployments make it easier to avoid data and analytic silos and should be considered where possible. However if this is not realistic many vendors have technologies to enable integration with other data sources.
As most organisations rely on a variety of data sources, it is important to have a unified data fabric that ties these sources together with a common query layer.
This creates a data virtualisation capability that can be used to easily incorporate additional sources of data, so new data sources can be linked to the existing sources in a common logical data model.
The use of data lakes and warehouses as centralised, well-governed repositories helps avoid data duplication and consolidates information from a variety of sources. In fact, around 60% of organisations said data lakes helped them gain a competitive advantage, one-third reporting increasing sales, and also lowering costs and improving the customer experience.
Choosing the right architecture
Data and analytics architecture should be consistent with the organisation’s overall information technology infrastructure and architecture and designed to enhance the exchange and use of data and analytics within the organisation.
It’s important that it’s scalable to match changing resource requirements and allows easy access and usage of data within the systems. With many organisations taking a hybrid cloud approach it’s important to include support for all cloud deployment options in your data architecture design. With the right design organisations can reduce management and administrative workloads and maintain the flexibility to easily re-platform as necessary.
Comment below to have your say on this story.
If you have a news story or tip-off, get in touch at firstname.lastname@example.org.
Sign up to the Government News newsletter