Cloud-Native Storage with Lightbits for Kubernetes Workloads

Explore deeper technical guidance

Cloud-Native Storage for Kubernetes

Performance, Persistence, Simplicity

Kubernetes
Storage,
Redefined

Kubernetes has become the platform of choice for modern, containerized applications. However, delivering high-performance, persistent storage that retains the portability and simplicity of Kubernetes is a major challenge—especially for stateful workloads.

This cloud-native storage solution is designed to deliver the performance of local NVMe flash, combined with the flexibility of disaggregated, software-defined storage, and the simplicity of standard TCP/IP networking. It allows developers and operators to run high-performance, stateful applications without compromising Kubernetes agility, resiliency, or efficiency.

Targeted Customers

Enterprises running stateful workloads in Kubernetes (e.g., MongoDB, MySQL, Kafka, Cassandra, Spark)

DevOps teams that need persistent volumes with flash-level performance

Platform engineers and SREs who want resilient, high-throughput storage without managing complex infrastructure

Cloud-native service providers looking for cost-efficient, software-defined storage for Kubernetes

Integrates Lightbits storage with Kubernetes via CSI and NVMe/TCP

High-Level Design & Components

Runs on standard x86 servers
Supports elastic RAID for drive failure protection
Achieves high performance and availability without custom NICs

Enables dynamic provisioning of persistent volumes
Transparent integration with Kubernetes StorageClass, PV, PVC

Standard TCP/IP transport
No need for RDMA, RoCE, or specialized NICs
Works with existing Ethernet switches and Linux kernels

Thin provisioning
Inline compression
Snapshots and clones for CI/CD and dev/test agility

Key Features and Capabilities

Our solution boosts performance, flexibility, and scalability. With technologies like Elastic RAID, Kubernetes integration, and storage flexibility, it optimizes infrastructure. Explore features that ensure seamless operations, fast recovery, and cost efficiency across hybrid and multi-cloud environments.

Local NVMe Performance, Network Storage Flexibility

✔️ Offers the same or better performance as direct-attached NVMe
✔️ No CPU/network overhead from pod migration or application replication
✔️ Utilizes standard 25–100 Gbps Ethernet

High Resiliency & Fast Recovery

✔️ Elastic RAID ensures rapid, local rebuilds without network saturation
✔️ Supports 2x and 3x replication across target nodes

Simple, Standards-Based Integration

✔️ Runs on any x86 platform with no proprietary components
✔️ Integrates via CSI, using familiar Kubernetes storage APIs
✔️ Fully compatible with major Kubernetes distributions (upstream, OpenShift, Tanzu, etc.)

Lower Total Cost of Ownership (TCO)

✔️ Increases flash utilization from 15–25% to much higher efficiency
✔️ Supports QLC NAND through intelligent wear-leveling and reduced write pressure
✔️ Delivers up to 10:1 data reduction with compression and thin provisioning

Developer-Ready Tools

✔️ Fast clones and snapshots for dynamic environments
✔️ Ideal for CI/CD, data science notebooks, and test environments