HPE Alletra delivers critical services for containerized databases like MongoDB, leveraging HPE CSI Driver for Kubernetes to containerize workloads while providing an on-demand as-a-service experience for instant access to data.
โBy Parissa Mohamadi, Product Manager for Storage Data Protection Solutions, HPE
Did you know that 96% of organizations are either using or evaluating Kubernetes[1] and that more stateful workloads such
This is where the underlying storage platform plays a critical role in containerizing workloads to meet these demands by providing:
- Feature-rich API delivering infrastructure as code so that developers meet the larger initiatives of storage infrastructure as a service
- Operations optimization enabling snapshots to protect container data, replication between on-premises systems, and multi-tenancy Kubernetes clusters across multiple storage systems
- Agile flexibility allowing for development of scalable, portable, and stateful apps for Kubernetes anywhere
The good news: HPE Alletra, the worldโs most intelligent storage for mission-critical apps, leverages the HPE CSI Driver for Kubernetes
In this way, HPE Alletra offers what it takes to containerize your workloads while providing an on-demand, as-a-service experience for instant access to data. For example, if you have stateful workloads running on HPE Alletra, you can containerize them to reap the benefits of:
- Resiliency and unprecedented simplicity
- Prediction and prevention of disruptions with app-aware resiliency
- Consolidation with low-latency performance
Take a database workload such as MongoDB. To containerize MongoDB, all you need to do is create a Kubernetes cluster first. Then, deploy the Kubernetes cluster using Rancher Kubernetes Engine (RKE) as the container management platform to simplify cluster installation and operations. Next comes HPE Alletra providing the storage infrastructure piece that delivers data services to MongoDB database such as snapshots, zero-copy clones, and volume expansion for test/dev purposes. MongoDB sharded or replica set workloads can be deployed using the MongoDB Enterprise Operator for Kubernetes and can be managed with MongoDB Operations Manager. Because data services can be automated, DevOps environments can be quickly deployed in a reproducible way. By leveraging the HPE CSI Driver for Kubernetes that is integrated into the Rancher Apps & Marketplace, HPE Alletra Container Storage Provider (CSP) provisions persistent volumes from HPE Alletra 9000 system as the back end.
Contrast this solution with one deployed in a local storage environment where interaction is required with the LVM on each host to capture snapshots instead of having these snapshots specified and captured at the Kubernetes level and having the operation off-loaded to the array. In this solution, snapshots of a complete and complex environment are created in one single operation and can also be used as a backup option. Cloning a MongoDB resource and utilizing snapshot groups can eliminate the time- consuming re-initialize operation that might be needed in a traditional backup/restore operation (propagating the database from primary to secondary).
Youโre also free of the burden of downtime to add and configure additional physical infrastructure when you want to scale your workload as this is done live and on an as-needed basis on your existing Kubernetes cluster.
To learn more, watch this demo video and read this technical white paper detailing the deployment of a stateful workload on Kubernetes and operational examples illustrating the benefits provided by HPE Alletra.
[1] Source: CNCF Annual Survey 2021
Meet Storage Experts blogger Parissa Mohamadi, Product Manager for Storage Data Protection Solutions, HPE
Storage Experts
Hewlett Packard Enterprise
twitter.com/HPE_Storage
linkedin.com/showcase/hpestorage/
hpe.com/storage
StorageExperts
Our team of Hewlett Packard Enterprise storage experts helps you to dive deep into relevant infrastructure topics.
