Platform Architecture / Silk DataPod

Silk DataPod: The High-Performance Data Layer for Your Cloud

A Silk DataPod turns standard cloud compute and storage into one high-performance data platform. Your databases connect to it just like native cloud block storage, with no code changes and no rearchitecting.

Silk's 3 Tier Architecture with Silk DataPod highlighted

Native Cloud Disks Tie Performance to Capacity

With native cloud block storage, getting more performance usually means buying a bigger disk, a premium tier or a larger VM, whether you need the capacity or not. A Silk DataPod separates the two, so you pay for the performance and capacity you actually use.

Native Cloud Storage

  • Performance tied to disk size and tier
  • Overprovision capacity to get the IOPS you need
  • Full copies consume capacity, time and budget

Silk DataPod

  • Performance and capacity scale independently
  • Up to 100× faster database performance
  • Up to 69% lower cloud costs
  • Zero-footprint snapshots and clones
  • No application or database changes

What's Inside a Silk DataPod

Every Silk DataPod combines a performance layer, a capacity layer and built-in data services.

Performance layer

Silk compute nodes serve I/O in parallel for consistent low latency and high throughput. Add nodes when demand rises and remove them when it drops.

Capacity layer

Data lives on cloud media managed by Silk. Capacity scales on its own, so you never overbuy one to get the other.

Inline data reduction

Compression, deduplication and thin provisioning shrink your data footprint, so you store and pay for less.

Zero-footprint snapshots and clones

Create writable copies of production instantly, with no extra capacity until data changes.

Silk Echo

Give dev/test, analytics and AI teams production-grade database copies that run alongside production without slowing it down.

Replication and recovery

Built-in replication and snapshots let you restore production data in minutes after an outage or ransomware attack.

How a Silk DataPod Serves Your Databases

Your applications see standard block volumes. The Silk DataPod does the heavy lifting underneath.

Your databases

SQL Server, Oracle, PostgreSQL

Silk volumes

Connect like native block storage

Silk DataPod compute nodes

Serve I/O in parallel

  • Data reduction
  • Snapshots
  • Clones
  • Replication

Cloud capacity

Media managed by Silk

Add compute nodes to raise performance on demand, and remove them when demand drops.

Why Teams Choose Silk DataPods

Performance you can scale, costs you can predict, and no changes to your applications.

Scale performance on demand

Raise throughput for peak periods or heavy batch windows, then scale back down, without migrating data or resizing disks.

Pay for what you use

Each Silk DataPod is priced on the capacity and throughput you provision, so you stop paying for idle premium storage and oversized VMs.

Runs in your cloud

Silk DataPods are deployed in your own AWS, Microsoft Azure, Google Cloud or AI cloud environment, so your data stays where it lives today.

Works with your databases

SQL Server, Oracle, PostgreSQL, analytics and AI workloads use Silk DataPod volumes exactly as they use native cloud disks.

Proven in production

In a commissioned Forrester Total Economic Impact study, the composite organization cut cloud storage costs on its Silk workloads by 50%.

See a Silk DataPod in Action

Bring one workload and see how a Silk DataPod changes its performance and its cloud bill.