The new cloud-native solution is built on the WekaFS file system, and integrated into unified storage solutions though reference architectures with object storage Technology Alliance partner vendors like AWS, Cloudian, Hitachi Vantara, IBM, Seagate, Quantum, and Scality.
WekaIO, which makes high performance scalable parallel file storage for data-intensive workloads, has unveiled a new cloud-native, storage solution that unifies and simplifies the data pipeline for accelerated DataOps. It creates a global namespace across private and public cloud ecosystems to facilitate data management for edge-to-core-to-cloud containerized applications, making it easy to manage petabytes of data in a single, unified namespace wherever it is in the pipeline. That, the company says, will deliver the top performance needed to accelerate AI and Machine Learning-based research like genomics and high-performance computing.
“What we are announcing here is around our data management capabilities, using a new cloud-native approach, and focused around our object storage ecosystem partners,” said Shailesh Manjrekar, Head of AI and Strategic Alliances at WekaIO.
The new solution falls within WekaIO’s strategy around DataOps – or Data Operations – for those tired of industry acronyms.
“We have three primary pillars in our strategy around DataOps,” Manjrekar stated. “First, we derive actionable intelligence out of that data set, through our performance and ability to cater to mixed workloads. Then we operationalize the data stores through this new way of doing data protection. Third, we also provide enterprise readiness and governance.”
The secret sauce here is the solution’s ability to extend the WekaFS namespace from high performance flash to an S3 REST-enabled cloud object storage system, to manage petascale datasets effectively without compromising performance.
“The way that we work with the object store is very unique,” Manjrekar said. “We provide a global namespace that keeps the metadata on the flash and uses the object store as a capacity tier. That’s a major differentiator.”
The efficiency comes from the fact that all the I/Os are serviced by the flash tier, while also being able to leverage the object tier for capacity scaling.
Manjrekar said that the way the global namespace works isn’t the only differentiator here. Weka’s instant and space-efficient snapshots capability makes traditional backup redundant, because their snap-to-object feature captures a point-in-time copy of the entire, unified file namespace, that can be presented as another file namespace instance in a private or public cloud.
“Snap-to-object captures everything as a snapshot and lets you push it to the object tier, and it snapshots metadata as well as data,” Manjrekar indicated. “Other solutions like NetApp require you to also have a backup solution like Commvault.”
Manjrekar also highlighted that the solution has end-to-end encryption that ensures data is always secure throughout its lifecycle.
“We work with a leading key management system, so you don’t need stealth encrypting drives,” he said. “That’s another major differentiator.”
Nine object store providers have partnered with WekaIO through Weka’s Technology Alliance Program to provide fully validated solutions.
“We spent three to four weeks with each partner on an interoperability test plan,” Manjrekar said
They include AWS S3, AWS Outposts, with the latter having certified WekaIo in September. Other ones include Cloudian HyperStore, Hitachi Content Platform, IBM Cloud Object System, Scality RING and Quantum ActiveScale, with the Quantum platform being the ActiveScale technology Quantum acquired from AMD in March of this year.
WekaIO has direct, channel and OEM routes to market, and Manjrekar said that all three would be used with this offering.