hamburger icon close icon
Linux on Azure

Linux on Azure Workload Migration: Challenges and Solutions

Read Next:

February 5, 2020

Topics: Azure NetApp Files Advanced9 minute read

With cloud adoption at the heart of digital transformation for enterprises, there’s been a lot of buzz around the feasibility of deploying and hosting enterprise business workloads in the cloud. With the majority of enterprise applications relying on a Linux operating environment, some workloads can be easily moved using the lift-and-shift approach, while others require careful assessment and planning prior to migration.

Microsoft Azure is a popular public cloud platform for enterprises, as it supports the majority of open-source technologies and applications natively. Azure NetApp Files (ANF) is a managed file storage service available in Microsoft Azure that helps address the requirements of Linux workloads in Azure require NFS file shares.

This blog will explore the benefits of ANF and what makes it the must-have storage solution for every migration plan involving Linux file workloads in Azure.

Azure Migration for Linux Workloads

When it comes to migration of Linux applications to Azure, namely tier-one applications like Oracle, SAP, and Hadoop, organizations are often apprehensive about the time and effort required, especially when file shares are involved. In addition to the complexity of doing so, other factors such as performance, scalability, compatibility, and high availability must be considered before migration and adoption of Linux applications.

Let’s delve into the challenges enterprises face when deploying Linux on Azure.

Performance Benchmarks of DB Workloads

Getting a quick response to database queries is a non-negotiable attribute when it comes to enterprise applications. Database platforms like Oracle and SQL use file shares extensively in their architecture, and file services in the cloud should be capable of sub-millisecond latencies and bare metal throughput.

Moreover, in an organization's DB landscape, different workload environments will have varying performance demands. For example, line of business (LOB) banking and retail applications require higher data operations (IOPS) in production systems, while QA and development databases can operate on file storage with less stringent IOPS requirements.

Scalability for Handling Big Data

Big data analytics applications handle large datasets, and the underlying file storage should be scalable in order to meet growing capacity demands. This often is inefficient with on premises deployments as it requires extensive capacity planning as well as upfront investment for storage devices.

Adopting cloud-based services helps to address this challenge, provided the file solution is capable of scaling up to the desired capacity on demand. The solution should also provide hassle-free Azure migration tools for moving data from existing systems to the target cloud file share.

DIY NFS File Shares

Building an NFS file server cluster in Azure with VMs and managed disks is one of the common solutions used during Linux migration for applications with NFS-share dependency. However, this approach requires far more effort when it comes to designing and planning the system, its high availability, security, and capacity management.

Maintaining and managing an NFS file service cluster may also require a dedicated team in the long run, which would be overkill considering the base requirement was just to make NFS file shares available for the migrated workloads.

Quick Deployment for DevOps

In the fast-changing, competitive IT landscape, an application’s success often depends on how fast an in-demand feature can be released to the market, with the goal being at least one day before your competition. This calls for a “fail fast” approach, whereby continuous integration and continuous delivery (CI/CD) environments can be deployed and destroyed on demand through DevOps processes, without causing storage overhead.

Heterogeneous Workloads

Workloads leveraging both Windows and Linux OS for different tiers is the norm in most data centers. With Linux workloads, there often is a combination of legacy applications that depend on NFSv3 and modern workloads using NFSv4.1.

When migrating such heterogeneous applications to the cloud, both NFS- and SMB-based file shares need to be deployed to support them. Manual configuration of these services only adds to the complexity of the migration and is something most organizations wish to avoid.

Cost-Performance Balance

Maintaining the right balance between cost and performance is an ongoing struggle, with the latter often compromised when optimizing the former. This results in lower performance of Linux workloads migrated to the cloud.

The right file service should be capable of providing different storage tiers to help achieve the ideal balance between cost and performance, without compromising on either aspect.

How Azure NetApp Files Can Help

ANF offers a managed file storage service in Azure that supports NFS and SMB protocols so that workloads migrated to Azure can benefit from enterprise-grade data management features without compromising performance. ANF is delivered as a Microsoft first-party service directly from the Azure portal. Built on NetApp’s trusted ONTAP storage solution and the Azure platform, the service offers high reliability.

Customers can access the service through through their existing enterprise agreements without the need for additional onboarding. It is also fully integrated with Azure management options, such as Rest API, Azure CLI, and PowerShell.

The service is highly available by design, allowing you to migrate your file-dependent Linux workloads with confidence and without having to worry about service uptimes. Added benefits include: multiple performance tiers, scalability, security, and data protection—among others—that help address the common challenges faced by organizations while migrating Linux workloads to Azure.

Multiple Storage Performance Tiers

ANF offers three service tiers—Standard, Premium, and Ultra—each with different performance benchmarks so that customers can select the most appropriate price/performance for their workload.

The Standard tier offers a throughput of 16 MB/s per provisioned terabytes, sufficient for test and development environments. The Premium tier, on the other hand, can be used for UAT or even production systems with throughput requirements of up to 64 MB/s per TB.

For mission-critical Linux applications with non-negotiable performance benchmarks, the Ultra tier provides a throughput of 128 MB/s per TB. With proven performance of over 300k IOPS, this tier is ideal for traction-intensive database workloads.

On-Demand Scaling and Data Migration

For applications handling big data, scalability is a critical feature. Each ANF volume is capable of scaling up to 100 TB to meet growing capacity needs. Moreover, the capacity doesn’t need to be provisioned upfront; the volumes can initially be provisioned as small as 100 GB and then later scaled to 100 TB as the data set size increases.

While migrating on-premises applications to Azure, customers can leverage NetApp’s Cloud Sync feature to move data sets to the cloud. Once data has been migrated, customers can then build advanced analytics solutions with Azure HDInsights with their data sets.

Managed Service Model

ANF eliminates the requirement of building complex NFS file server clusters in Azure. With just a few clicks, the required file shares can be deployed and attached to the application workloads. This reduces deployment and onboarding time from days or weeks to a matter of minutes, thereby accelerating the migration process.

Faster Time to Market

Multiple copies of ANF volumes can be created by taking point-in-time snapshots of the data, with a minimal storage footprint and without harming source data integrity. Snapshots along with instant copies of CI/CD environments ensure more frequent and efficient application feature releases. The reduced time to market (TTM) gives organizations a competitive edge.

Multi-Protocol Support

For environments with heterogeneous workloads—Windows and Linux—ANF offers both NFS- and SMB-protocol support. With both NFSv3 and NFSv4.1 supported, migrating legacy as well as modern Linux applications to Azure is simplified. .

Cost Optimization

ANF is charged per GB per provisioned capacity on an hourly basis. The Standard tier offers the lowest price, while the costs for the Premium and Ultra tiers are higher. Organizations have the flexibility to choose the most economical service tier for their needs without compromising performance. For example, applications capable of operating at a throughput of 16 MB/s would benefit most from the Standard tier in order to achieve the optimal balance between cost and performance.

Successful Linux on Azure Migrations  with Azure NetApp Files

When deploying Linux on Azure, it’s important first to examine all application components carefully and then to lay out a detailed migration plan. Planning Linux application migrations that have a file requirement means looking to a Linux file service such as Azure NetApp Files (ANF).  ANF services are closely aligned with Azure design best practices as ANF uses native Azure resources like VNet and RBAC and requires no additional product expertise or learning curve.

Here are a few useful tips for migrating workloads to Azure with Azure NetApp Files:

Capacity planning: When planning your applications capacity requirements, build capacity pools that provide a buffer for your day to day capacity spikes.ANF volumes are created within the capacity pool and you don’t want to run into capacity limits unless its planned.

Network access: ANF is deployed in a delegated subnet which means that your data is within your VNET security domain and not on an open IP address. With only resources connected to the VNET directly or via peering/VPN/ExpressRoute being able to access the service, proper connectivity of workloads is required to avoid network-related issues.

Access control: Azure RBAC should be configured to ensure that only authorized personnel can create/manage/delete file shares. Creating NFS export policies is also advisable so that only authorized clients can access the service.

Performance optimization: ANF performance is calculated as throughput per provisioned TB such that the volume’s performance can be dynamically adjusted to improve the throughput offered by only adjusting the capacity. This helps fine tune the performance of migrated applications on demand.

Cloud Adoption Done Right

ANF provides best-in-class support for Linux workloads on Azure, addressing the major migration-related pain points—performance, scalability, agility, and security. Advanced features such as instant copies, multi-protocol support, and data encryption make it the perfect choice for file-share-dependent workloads.

Migrate your file-share-dependent workloads with ease. Subscribe to Azure NetApp Files today.

Related Linux on Azure content:

Cloud Data Services

-