VMworld: JetStream Data Protection Software With Accelerate and Migrate, as Part Cross-Cloud Data Management Platform

JetStream Software Inc. announced JetStream Data Protection, enabling cloud service providers a solution to deliver data availability services to enterprise customers.

Click to enlarge

The Data Protection software joins the available solutions, Accelerate and Migrate, as part of the Cross-Cloud Data Management Platform.

The company will be exhibiting at VMworld 2018 in the ‘New Innovators Pavilion’, August 26-29, 2018, in Las Vegas, NV.

“Data protection is the most common use case for enterprise hybrid cloud,” explained Tom Critser, co-founder and CEO, JetStream. “So it is no surprise that the market is moving away from legacy snapshot-based replication architectures and adopting continuous replication methods for cloud-based data resiliency.“

Advanced design for data protection services
The Data Protection software provides a spectrum of data protection capabilities for an on-premises VMware, Inc.‘s environment, including:

Non-disruptive continual data replication to the cloud for recovery and failover
VM failover to a cloud-based BC service
VM backup to cloud-based NFS stores and/or object stores
Policy-based availability for cost-efficient SLA management per recovery group
Low-latency logging on non-volatile memory

Click to enlarge

Continuous data capture through IO filters
Many traditional backup technologies still capture data in intermittent snapshots, which introduces latency in application performance and limits the minimization of RPO. As an alternative to snapshots, some technologies capture data I/O through software agents or virtual appliances, methods that introduce a number of performance and compatibility problems. Continuously capturing data through an IO Filter does not suffer those drawbacks.

In VMware environments, the company’s solution enables continuous data capture through the vSphere APIs for IO Filtering (VAIO), ensuring compatibility with the entire VMware environment, including any type of on-premises hardware such as SAN, NAS, Virtual SAN or other hyperconverged infrastructure (HCI). Because it employs the VAIO framework for continuous data capture, the solution doesn’t require software agents in the VMs, doesn’t require a virtual appliance in the data path, and is a supported VMware Ready solution.

The firm’s engineering team collaborated with VMware as the co-design partner for the IO Filter API framework. The company has already released two products based on the IO filter APIs: Accelerate for cloud resource and performance optimization and Migrate for live migration of VMs to the cloud.

Click to enlarge

Cross-cloud data protection
The key components of the firm’s Data Protection software include:

IO filter: the software component deployed to each VMware host for continuous data capture for every VM in the protected recovery group
Data availability policy engine: management component that defines data replication and instance recovery policies for each recovery group
Low latency replication log: mechanism for maintaining the replication log, which can be deployed on an available non-volatile memory appliance for minimal impact on application performance
Cloud store adapter: software that pre-processes data for backup to a specific object store or NFS store
Crash-consistent data pipe: high-throughput, multi-channel software connection that provides a crash-consistent transfer of data from the on-premises data center to the cloud destination, ensuring continuous data coherence
Cloud acceleration tier: low latency data tier in the cloud that enables both zero RPO with data consistency and efficient use of cloud resources

Click to enlarge

Why is architecture important?
Serge Shats, co-founder and CTO, JetStream, described the reason for the design of the Data Protection software: “We have the ability to continuously capture data for recovery in real time, as it’s being written to storage, but that’s just the start. Beyond capturing data through the IO Filter, we have to satisfy a considerable number of objectives.“

Objectives include:

No degradation of application performance on premises
High scalability to support environments with thousands of VMs
Low latency replication logging for scalable continuous data replication
Support for multiple, concurrent availability policies, assigned per recovery group
Replication to the recovery destination most appropriate for each policy, from a warm VMware environment for cloud failover to an economical object store backup site
Crash-consistent connectivity with each recovery destination
Multiple recovery options: e.g., point-in-time (CDP) or whole group failover (DR)

There are hundreds, sometimes thousands, of VMs across the data center, and they’re constantly writing data.

“It’s a firehose of data I/O,” explained Shats. “If the connection to the cloud, relatively speaking, looks more like a garden hose, you have a problem. So to provide a highly scalable solution for data replication to the cloud, we employ 1) a low-latency environment for the replication log on premises; 2) the multi-channel, high-throughput data pipe to transport data to the cloud; and 3) a high-performance data acceleration tier in the cloud recovery destination.“

JetStream Accelerate and Migrate are available. The Cross-Cloud Data Protection software is in technology preview. All firm’s products are distributed via enterprise technology resellers, systems integrators, cloud service providers and MSPs.

Additional resources:
Blog: JetStream Migrate Now Available
Blog: JetStream Accelerate 4.5

Read also:
JetStream Software: Cross-Cloud Data Management Platform With Cloud Migration
For workload migration, resource elasticity and BC across multi-cloud infrastructures
2018.04.12 | Press Release