Dataware for data-driven transformation

Five Challenges You Will Face When Going All in On Cloud

Contributed by

10 min read

The move to the cloud is imminent for many organizations. Increasingly, enterprises are leveraging cloud services to increase business and IT agility and modernize outdated data centers. In the process, they are forced to migrate apps to the cloud or even replace the apps altogether. One of the significant driving forces behind this trend is the explosion of data-intensive apps as well as infrastructure migrations to the cloud as part of an organization's digital transformation.

Before migrating their analytics workloads to the cloud, IT departments need to fully understand the complexities, know what to move, and know how to move their data and applications without impacting business operations. If you're making the leap to the cloud, here are some of the biggest challenges you might have to overcome in order to benefit from the advantages of infrastructure as a service.

Five Challenges

1. Planning and Executing a Smooth Cloud Migration

Despite the cloud's rapid rise, there's still a general lack of understanding as to why moving to the cloud can be beneficial. There is also a lack of technical expertise in transferring workloads to the cloud. Before you start to build out your cloud, hybrid, or multi-cloud environment, you first need to understand how cloud infrastructure will impact your data access and all related applications. You can't move your apps all together, at the same time. Most organizations have thousands of applications; IT must understand or relearn and often rewrite them before moving any of them to the cloud. The most common example is an app that previously stored data in a NAS and now has to speak a new protocol to speak to object storage. With MapR, that same application doesn't need to be rewritten, because the same NAS interface can be provided in cloud, while a strategic portion of the underlying data will still get the cost advantages of object storage.

Additionally, some data and workloads may need to remain on-premises. It could be for security or compliance reasons. Or, when you already have an on-premises data management and processing solution such as the MapR Data Platform, you may want to expand its capacity for storing data and running new analytics in the cloud rather than adding hardware and other resources to your on-prem infrastructure. The MapR cloud-aware provisioning and management tools allow you to use many existing or new cloud infrastructure-specific features without the cost and complexity of customizing cloud deployments on your own.

As you can see, a successful migration to cloud requires proper planning and the right technologies.

2. Vendor Lock-In

The issue of vendor lock-in boils down to the lack of standardization in how data is stored, secured, and moved between cloud providers. Furthermore, incompatibility between comparable components and services across clouds forces companies to choose a single cloud, making it impossible to take the best-of-breed approach with the best properties of each provider. Then, the centralized focus of most cloud providers prevents companies from adopting hybrid architectures to more easily move away from their legacy infrastructure and embrace a distributed edge architecture.

MapR works everywhere. We give you the freedom and power to work, however you like, wherever you like. If you are already familiar with running MapR on-premises, you will find that your knowledge and practices carry over to the cloud. One significant factor for using multiple clouds is that some workloads run better on one cloud platform, while other workloads achieve higher performance and better cost-efficiency on another platform. To better prepare for disaster scenarios, we advocate for a multi-cloud, multi-data center strategy. Data stored in the cloud is susceptible to system-wide outages. One recent example involved the failure of AWS, which took down a big part of the internet, costing $150 million to companies depending on Amazon's service.

With MapR, you can replicate everything from the first data center to the second and load balance across the different cloud providers. This means you have a solid disaster recovery and business continuity plan in place. You can choose which workloads to run in which places, based on cost. You're not locked in to any specific cloud provider. You have options.

MapR has an open API architecture, which assures interoperability and avoids lock-in. All of the APIs are standard and accepted by the industry. This allows you to easily build software against our platform in any environment. With the MapR Data Platform available in the AWS marketplace and Azure marketplace, you can have MapR clusters up and running much faster and with much less effort on clouds.

3. Controlling Costs

Migrating workloads to the cloud brings significant benefits, but there are also significant risks if you lose sight of spending on cloud resources. Getting data into a public cloud can be easy and cheap, but getting that data back out can be prohibitively expensive. Cloud services can lead to zombie services that have little or no utilization if services are not managed in a timely manner. It only gets harder when you have to manage multiple clouds.

One way to reduce spending on cloud resources is to use object tiering to optimize resources on-premises or off-premises. Object tiering provides policy-based, automated tiering functionality that lets you seamlessly integrate and move inactive (cold) data to and from cloud object storage. By moving cold data that is rarely accessed to cloud storage, you can retain it at minimal cost. Additionally, you can eliminate the need for archiving software and cloud gateways that add additional complexity.

MapR object tiering (roadmap item) allows for data to be transparently offloaded to cloud object storage and accessed as if they were locally present. We maintain consistency by automatically retrieving and sending any updated file data back to the cloud so that the cloud always contains the latest version. This allows companies to get the best of both worlds – real-time performance on recent, or "hot" data, and low cost on older, "cold" data – without having to actively manage and move data and without applications being aware and written for different data systems.

4. Data Sprawl

As companies continue to move more workloads to the cloud, data sets are also being continuously created, managed, and accessed. When apps and data are spread across multiple cloud services, you end up with silos of data and analytics, which impede productivity and impact data integrity. You need a data strategy and the data architecture to address the challenges of sprawling data, especially in today's multi- and hybrid cloud environments.

When data lives outside of your on-premises data center, you need to think beyond consolidating data in one place. It's more about connecting data and analytics across multiple environments. The first step toward reducing the risk of data sprawl is managing all your data in the MapR Data Platform. The MapR global namespace allows you to run clusters worldwide and provides a consolidated view into files that are in different physical locations. That means applications can access data in any cluster seamlessly without having to worry about where the data is. If an application is using data that has been aged out to the cloud, the application doesn't need to know that; it still just runs the query. The beauty of this is that you eliminate data silos, enabling data to flow back and forth between MapR clusters across clouds and between MapR and existing enterprise systems.

5. Security and Compliance

Maintaining security is a shared responsibility between you and your cloud provider. Although major cloud providers understand the importance of security in the cloud and are compliant with HIPAA, PCIDSS, ISO 27001, and more recently, the EU Data Protection ruling, organizations need to see to it on their own by using a data platform with security features built-in. That means you need a platform that has encryption, access control, risk management, and other security features unified across all the different kinds of applications and data to ensure consistencies in policies.

This is what the MapR Data Platform is all about. We provide a very rich set of platform services, with global high availability, global namespace, and unified data protection, where we can roll forward or roll back your data at any time. It's real-time and fully multi-tenant to securely support multiple organizations with different users and groups.

Cloud migration is a complex and critical process. Selecting the right data platform is necessary to ensure success in the cloud. If you are still struggling with these cloud challenges, then talk to us. We can help you address your concerns.


This blog post was published May 16, 2018.
Categories

50,000+ of the smartest have already joined!

Stay ahead of the bleeding edge...get the best of Big Data in your inbox.


Get our latest posts in your inbox

Subscribe Now