Pure to offer on-prem object storage as Snowflake data source

Pure Storage has announced a partnership with Snowflake that will allow customers to integrate the cloud data warehouse platform with their on-premise fast object storage from the storage supplier.

“What we’re hearing from customers is that they want to access the power of Snowflake while keeping data on disk, on-premise, for reasons of compliance, security or performance,” said Robert Lee, chief technology officer at Pure Storage.

Snowflake provides cloud-based data warehouse functionality, in which customers can point the platform at cloud-resident data stores, but now also integrate it with on-premise Pure Storage arrays.

Snowflake provides SQL-based querying to multiple data sources – including those not necessarily in SQL databases – to allow for analytics operations to be run.

Under the Pure and Snowflake partnership, customers will be able to run directly from local data without having to upload it to the cloud. Benefits of this, according to Lee, include speeded up time-to-market and time-to-value, along with the ability to use the data in different applications without needing a second copy, and so simplifying workflows.

Lee said Snowflake brings uniqueness in terms of being a cloud-based service and an architecture that allows for independent scaling of compute and storage.

“We’re partnering with a market leader that uses SQL as an access mechanism to much broader sets of data, such as log and IoT [internet of things] data, and can pull structure from it.”

The move follows a similar announcement earlier in May from Dell EMC announcing a partnership that would see Snowflake integrate with its on-premise object storage.

While Snowflake works with public cloud storage such as Amazon Web Services (AWS), there’s no intention to integrate it with Pure’s Cloud Block Store offering, largely because the focus is object storage, said Lee.

“This is an object storage-focused integration – which is how most customers are collecting this type of data – so much more focused on FlashBlade object storage,” he said.

“Cloud Block Store is more focused on block storage for traditional applications, and Portworx focused on providing native container storage, which are both key parts of our public cloud portfolio.

“Where we see more applicability for this integration is in the hybrid cloud portfolio and customers connecting cloud-applications to on-premise storage, which is where FlashBlade stands out,” added Lee.

FlashBlade is Pure Storage’s fast file and object family, which unifies file access and object storage in a single array to provide rapid access to large-capacity storage. It is aimed at what have traditionally been secondary storage use cases – backup, archive, analytics datastores – but which have evolved to require rapid input/output for analysis, recovery, and so on.

Snowflake competitors include Databricks and AWS Redshift in the cloud, and Vertica on-premise.

The fruits of the Pure-Snowflake partnership will be in public preview by the second half of 2022.

“Our joint solution with Snowflake reaffirms our commitment to simplifying how organisations interact with data to extract the most value and drive meaningful business outcomes,” said Lee.