Tech Insights
cancel
Showing results for 
Search instead for 
Did you mean: 

Why protecting AI containers is important

Supporting containers with data backup and mobility at scale is a critical need for AI and machine learning environments. Learn the top requirements for protecting containerized applications.

Container technology adoption is increasing at an amazing rate. The global application container market size was valued at AI containers-blog.jpgUSD 1.5 billion in 2018 and is expected to reach USD 8.2 billion by 2025, according to a new report by Grand View Research, Inc. and other analytics firms, registering a 26.5% CAGR during the forecast period.

This growth is the result of how easily containerized applications port and deploy across different environments. Indeed, containers virtually package the applications with everything it needs to run with (configuration files, dependencies, etc.) and isolate them for the deployment environment. This enables containerized applications to easily run on different environments such as local desktops, virtual and physical servers, development, testing and production environments, and private or public clouds.

In addition to portability, containers have another big benefit: a physical server can host more containers than virtual machines. Containers share access to an operating system kernel, so they require much less space than virtual machines because each container shares the host's OS. Indeed, the average size of a container is within the range of tens of megabytes while virtual machines can be up to gigabytes in size. Hardware efficiency is another benefit driving market adoption.

Another factor driving container growth is their increasing popularity for deploying artificial intelligence (AI) and analytics applications. Containers are becoming the standard way to build and deploy machine learning (ML) models, creating real-time analytics pipelines and running batch analytics and ETL jobs. Their portability across different environments makes containers the perfect vehicle to manage the full lifecycle of AI/ML models and, for that matter, most any analytics application. 

The massive adoption of containers for analytics and AI/ML applications is creating a demand for containers for stateful. Analytics and AI/ML applications are generally stateful applications which use and generate a lot data and as a result, need persistent storage.

Three good reasons why containers need data backup too

Even if we can’t predict that containers will completely replace virtual machines (VMs), we are sure they will play an important role in the future of IT, driving the need for the same levels of protection and disaster recovery (DR) requirements in place for servers and VMs.

More specifically, consider these three very good reasons why data backup is needed for container environments such as Kubernetes or Docker and their associated applications:

  1. Recover containerized applications from failures and disasters
  2. Replicate the environment for migrating a test/dev environment to production, or from production to staging before an upgrade
  3. Migrate container clusters more easily

What are top requirements for container data protection?

Without a doubt, supporting containers with data backup and mobility at scale is a critical need for AI/ML environments. Keep these key requirements for protecting containerized applications in mind:

Implement seamless operations and policies across on-premises and clouds.

Containers can span multiple environments both on premises and in the cloud or even in multi-cloud environments. Operational simplicity for container deployment and data management policies that can span across multiple environments on premises and in cloud are strongly needed.

Backup and restore at the application level, not at VM/server level.

By nature, containers are not bundled with physical servers or virtual machines. One of the key challenges for protecting them is managing this dynamic deployment. The target of container protection is the quick restore of the full operation ability of applications rather than the backup of their data or configurations.

Remember automation is key.

Enterprises want to quickly deploy in production AI and analytics containerized applications and pipelines and they need to quickly and securely restore these applications. Fully automatic application restore is the key aspect of the solution. 

Keep security, security, and security top of mind.

Security in not an option. Protection could not be only at data level (e.g. encryption) but also at the application and user level (e.g. role-based access control or RBAC) and multi-tenants too.

Enterprise-grade container platforms are here to stay

 Container adoption is growing at a continuingly fast rate, with usage expanding over traditional stateless models. Indeed, stateful containers are becoming standard way to manage AI applications. Unfortunately, containers haven’t been designed to be protected from failure—and that’s why data backup is not so easy. Stay connected on this blog channel to learn about how HPE can protect your container platform.    


Andrea Fabrizi
Hewlett Packard Enterprise

twitter.com/HPE_Storage
linkedin.com/showcase/hpestorage/
hpe.com/storage

twitter.com/HPE_AI
linkedin.com/showcase/hpe-ai/
hpe.com/us/en/solutions/artificial-intelligence.html


Related articles

0 Kudos
About the Author

AndreaFabrizi1

Andrea is Senior Product Manager for Big Data and Analytics Solutions at HPE.

Events
Starting June 23
HPE Discover Virtual Experience
Joins us for HPE Discover Virtual Experience live and on-demand
Read more
Online Expert Days - 2020
Visit this forum and get the schedules for online Expert Days where you can talk to HPE product experts, R&D and support team members and get answers...
Read more
View all