This major series of enhancements to Pure Storage’s Kubernetes platform expands the number of use cases in which it can be deployed, and makes it easier to deploy it.
Venkat Ramakrishnan, VP of Products and Engineering at Portworx
Today, Pure Storage is announcing a significant series of enhancements to their Portworx Kubernetes portfolio. These include general availability of an enhanced version of Portworx Data Services, their Kubernetes Database as-a-Service Platform. They also include the 2.10 version of their Portworx Enterprise data management platform, which has significant improvements, and which the company said has been very popular in previews.
“This is a major upgrade of the portfolio,” said Venkat Ramakrishnan, who is VP of Products and Engineering at Portworx, and has been there since 2016, years before Pure acquired them.
“Portworx will always be the category creator in this space,” Ramakrishnan said. “We pioneered data management and data protection for Kubernetes, and with the new data services, we are creating new categories of data service for Kubernetes. We are the thought leader and the market leader in this space, and we have always been 12-18 months ahead of our competition.”
Ramakrishnan said that the new announcements should extend that lead.
“We are announcing the general availability of Portworx Data Services,” he stated. “That is the number one announcement that we are making. These data services will run any database anywhere in any Kubernetes cluster on any cloud. “We automate and run the cloud hosted control plane on any of the databases.”
The Portworx Data Services platform can be used to expedite deployment, Day 2 operations, and protection – for any data service – with a single click. It this removes the complexity that inherently stems from the use of different types of data services and different databases. Most popular data services, including Cassandra, Kafka, PostgreSQL, RabbitMQ, Redis and Zookeeper are supported now, with more services coming soon.
The platform is not completely new, but its pricing and data consumption models have been revamped, Ramakrishnan said.
“We have offered it this last couple of years, but we have aligned our Go-to-Market and pricing, and there are some differences from other Pure storage consumptions,” Ramakrishnan noted. “Other Pure models look at TB hours and capacity utilization usage. We bill by the number of node hours. For Backup node, hours are also how its done. Portworx data services are based on the workload hours. – we have always tried to look for the simplest pricing strategy. Our goal on pricing is to be as simple as possible for our customers.”
Pure is also announcing the general availability of Portworx Backup-as-a-Service on AWS. It supports backup and restore of any Kubernetes services including Amazon EKS, Red Hat OpenShift, Google Kubernetes Engine, VMware Tanzu, and Microsoft Azure Kubernetes Service
“With this launch of Data Protection for Kubernetes, we are launching data protection for our backup product as a service,” Ramakrishnan said. “It will be available as a SaaS service. Previously, customers had to download it and install it on their clusters. We have completely SaaSified it.”
Ransomware protection for Kubernetes has also been enhanced on backup as a SaaS and with Portworx 2.2.
“We have always offered certain guarantees for product that prevented many malicious attacks,” Ramakrishnan said. “We have secured data at rest and in flight – so that it never leaves. We have already supported storage multitenancy, with a fence to determine who can access their volumes. Now object locking comes into play if someone takes away infrastructure, to fully secure ransomware protection and provide encrypted multitenanted, role-based access control.”
Finally, Portworx Enterprise, version 2.10, was announced.
“Portworx Enterprise 2.10 has a lot of nifty features,” Ramakrishnan said. The key here is support for five new use cases for workloads in Kubernetes production environments.
One is multi-tenant platform-as-a-service with application I/O control.
“This lets you really go into large multi-tenant infrastructure with hundreds of apps, because the I/O control lets Kubernetes stop the apps from stepping on each other,” Ramakrishnan indicated. “We can now improve density when they implement more apps.
“The other big thing is IPv6 support for Kubernetes applications in a 5G/edge environment,” Ramakrishnan indicated. “Most IoT edge deployments run on IPv6. It will be ready in June.”
Another new feature uses the ability to lift and shift legacy applications with Kubernetes virtualization. This allows end-to-end support for KubeVirt and Openshift virtualization in Kubernetes managed platforms.
“We also added a unified storage overlay capability with Object Store Service, to provide a unified storage overlay solution with file, block, and object store services,” Ramakrishnan said. “In addition, we increased storage and performance by making them more application aware, to place the data closer to the applications.”
A hot data tier backed by high performance storage with PX-Fast allows customers to take advantage of high IOPS and low latency of any high performance storage backend infrastructure, including NVMe, without compromising the advantages of Portworx Enterprise, if they want that.
“Another nifty feature in Portworx Enterprise 2.10 is a volume trash can or recycle bin for Kubernetes volumes,” Ramakrishnan indicated. “This is for people who accidentally delete things. This lets them choose things to be archived so they can restore something if they delete it by mistake.”
Ramakrishnan also gave some previews of the 2.11 version due in June.
“It will provide direct access for Flash Arrays and Flash Blade,” he said. “We also previewed Portworx FAS, which has close to bare metal speeds, and will be a high performance offering for HPC computing. We will show it this week at KubeCon. It will be in a June tech preview and will go live in October.”
