Gary Lyng
Vice President Product and Solutions Marketing, Hitachi Vantara
October 04, 2023
Finding meaningful cost savings is a high priority for companies running their enterprise architecture in the public cloud. Cloud costs have ballooned in recent years and account for over 30% of IT budgets, reports IDC.
Enterprises have sought to shift their cloud workloads from on-premises to public clouds, depending upon pricing and performance, to provide themselves with better leverage and business flexibility. But database migration costs — spiraling well into the thousands per migration — have scuttled their expected hybrid cloud savings.
Some companies decide not to “lift and shift” all their workloads, choosing to optimize some of them them for higher performance in the cloud while keeping some on-prem. As a result, cloud migrations often engage consulting services that do refactoring — rearchitecting on-premises workloads to optimize them for a platform-as-a-service (PaaS) model.
Some companies tap third-party service providers because they need more in-house resources with the time, tools, or skills to manage their databases. These firms also face other operations challenges, such as determining if their data is updated or properly compliant from a privacy and security standpoint.
Yet most cloud migrations don’t require elaborate processes or extraordinary expenditures. The most sustainable and effective means of optimizing data migrations start with better metadata –a central nervous system for cloud migrations. With that foundation – a form of data modernization — cloud operational costs can be better contained with a combination of automation and self-service migration. Once an organization masters cloud data migration, it can utilize hybrid cloud tiering to balance workload costs and improve performance.
What steps must companies take to gain the upper hand in their hybrid cloud data migrations?
Data flows into clouds from a myriad of applications, often in different formats and from multiple locations. While many organizations attempt self-service migration and cloud data management, some data teams struggle to improve their efficiency and eliminate redundancies or manual validation steps.
The modern approach to data integration applies business glossary and machine learning capabilities to automate data discovery. Using AI, a modern data catalog can validate business rules and quickly assess critical quality metrics across the enterprise. The AI enables the data catalog to classify structured, semi, and unstructured data quickly.
Automating the cataloging process can accelerate searchability and improve time to insights, all while satisfying rigorous compliance standards. Whether the process is fully automated or involves some degree of DataOps curation, the results must be comprehensive. The last thing a data team wants to do is migrate an incomplete data set to Snowflake, MongoDB, AWS, Azure or GCP.
Automation can enable significant flexibility starting at the data pipeline level. With adequately managed data pipelines, firms can gain the ability to move data wherever workloads are needed — from on-premises to the hybrid cloud. Hybrid clouds can deliver outstanding performance, and business flexibility, especially with containers that enable shifting workloads and data flows that can be automated without manual intervention.
Ultimately, automation can reduce unnecessary cloud expenditures and enable organizations to optimize storage and provisioning costs. Hybrid clouds can only deliver significant value if data teams can shift and run data workloads on the most cost-effective platform. A data calalog can help. To learn more about this subject and see a solution from Hitachi Vantara, please visit this site.
Gary Lyng is Vice President, Product Solutions Marketing, Hitachi Vantara.
