In this video, we will look at the use cases, considerations, and methods for migrating enterprise data to a cloud environment. What is cloud data migration? Simply, it is the process of moving on-premises data to the public cloud or from one cloud to another. At the core of every application is data. If you are migrating your application to the cloud and adopting a cloud native approach, you will need to consider where your data needs to be located, secured, accessed, and how to cost effectively migrate and host it. The case for cloud data migration is compelling, especially if the data is being used by an application already migrated to the cloud. Many of the same use cases and benefits discussed for cloud application migration also apply to data migration. Some of these use cases include being able to: Benefit from cloud’s dynamic storage infrastructure to scale capacity with simplified provisioning. Assure application critical data is highly available and co-located with the cloud-native applications that depend upon it. Provide automated Backup & Disaster Recovery through public cloud provider multi-region capabilities. Improve content delivery speeds for customers in different geographies; perhaps creating secure, cloud-based Content Delivery Networks (CDNs) on edge servers. And taking advantage of API gateway features to secure and govern access to the data for different consumers and business roles. As you can see from these use cases, the choice to intelligently select cloud environments to host eligible data workloads is increasingly desirable and can help maximize application value, efficiency, and accessibility. With data migration to cloud, there are many considerations, sometimes even more than with application migration. Repeatedly, surveys have found that enterprises underestimated the planning needed for cloud data migration. Let’s examine some of the considerations that impact data migration choices; the first being the “Type of Workload”. Many specialized workloads such as databases, system backups, and virtual machines may need special tooling to migrate. If this is the case, you will need to confirm your target cloud provider is able restore and support them in their environment. Likely, the most critical thing to consider is factoring in the timeframe the data needs to be accessible to your customers. The primary dimension is the amount of data that needs to be transferred and weighed against the transfer methods available. Work with your chosen cloud provider to understand the transfer options, which include over-the-network and physical transport of storage devices. If networks are a possibility for transfer, actual speed and reliability need to be factored in before finalizing any decisions. For physical transfers, transport logistics becomes a factor. Most importantly, whenever you relocate data to a cloud provider’s datacenter a thorough review of the security and compliance requirements around the various classes of data needs to be performed. For example, companies should verify the cloud provider has recognized security and compliance certifications, such as ISO 27017 for Cloud Service Security or Service Organization Controls (or SOC) standards, which can go a long way in establishing trust in assuming off-premises governance risks. Beyond general certifications, providers must also adequately address strict legal requirements governments impose on the handling of classes of data in regulated industries such as banking, finance, and healthcare. Plus, the provider must adhere to ever-evolving governmental privacy laws which dictate the handling of all customer “Personally Identifying Information” (or PII). Often, this data must be stored and managed in the same country or governing entity in which the customer resides regardless of industry. Examples include: the General Data Protection Regulation (GDPR) for companies operating in the European Union (EU) And government data protected by the Federal Information Security and Management Act (or FISMA) in the U.S. Finally, data transfers must be secure via strong encryption methods while “in motion” over networks and “at rest” in storage devices. Auditable documentation that tracks “chain-of-custody” as data transitions from origin device in the target device should also be provided. This includes being able to identify uniquely any intermediate transfer devices, vehicles as well as the persons who have custody of these devices at each point. Once the transfer is complete, proof that all copies of the transient data have been erased should also be provided. The best option for your specific data migration project depends upon how much data you need to move, how quickly the migration must be accomplished, the types of workloads involved, and your security requirements. There are a few options to migrate your enterprise-local data to a cloud data center; they fall under two broad categories: “online” or network transfer and “offline” transfer using physical devices. The Online transfer method moves data from the source location over Wide Area Networks (WAN) to the target cloud provider. If a standard internet connection is not deemed reliable, secure, or fast enough, providers usually have dedicated, high-speed network alternatives for secure transfers in many major cities. If expense is still an issue, it might be worth exploring if transferring during “off peak-usage hours” can reduce costs. Off-hours transfers may also have the side-effect of increasing the speed and reliability of the transfer as the impact from others using the same bandwidth would be reduced. By contrast, offline transfer methods involve storage appliances that are physically shipped between an origin, “source” location and a target cloud provider. While, in theory, you could use online migration for any amount of data, time constraints and cost limit its feasibility when dealing with large amounts of data. For example, if you need to transfer 100 Terabytes of data over a network at a speed of 100 Megabits-per-second, it will take nearly 107 days. However, with a reliable network that supports 10 Gigabits-per-second, the transfer would take just over 24 hours. For transfers involving larger amounts of data in the petabytes range or want more determinism of timeframe, an “offline” option using physical devices is likely the only realistic option. Some cloud providers may allow you to “provide your own device” if it is supported by their physical datacenter infrastructure and management software. Additionally, one of the most convenient and often most affordable options is to have the cloud provider ship you a specialized device for fast, local migration. Once ready, the device can be shipped back for installation or transfer into the target cloud datacenter. While planning to transfer data, there are some other best practices that can be applied. Verify projected transfer timelines against business requirements and impacts. Identify data classes that may cause issues with format and need transformation or require special security attention. Plan to transfer data for mission-critical workloads in stages, testing at intervals throughout the process that the migrated data maintained its integrity. During migration, if data needs to be transformed, perhaps explore the use of cloud native storage options such as NoSQL databases and Cloud Object Storage (COS) for higher efficiency and feature accessibility. You should now be familiar with: Use cases for cloud data migration, Considerations and best practices for data migration, and Common transfer options.
