Block Storage Demystified – 10 Most Asked Questions from Storage Managers

Question 1: Why Lightbits block storage and not other alternatives?
Compared to alternatives with proprietary hardware, with Lightbits block storage, you enable the following that others don’t:

• Lower Total Cost of Ownership (TCO). Deploy on your choice of commodity, off-the-shelf servers, freeing you from vendor lock-in, enabling better hardware pricing, and managing your hardware lifecycle independently.
• True disaggregation. You can scale your OpenShift/OpenStack compute cluster completely independently of the storage cluster—freedom to add storage or compute servers to your clusters.
• Cloud-native. Designed for modern cloud environments like Kubernetes and OpenStack. The architecture, API, and CSI/Cinder integrations are purpose-built for dynamic provisioning and automation in popular orchestration platforms.

Question 2: Compared to other software-based alternatives with Lightbits, you enable the following, with others you don’t:
Ground-up design for large-scale, disaggregated storage using NVMe over TCP (NMVe/TCP). Lightbits creates a cluster-wide pool of NVMe resources that can be accessed by many clients with minimal CPU overhead. This enables scaling out as a large, centralized pool of storage. You can add more storage nodes to increase capacity and performance for all clients simultaneously.

Lightbits offers a rich set of enterprise data services across its entire storage pool, including intelligent flash management to increase SSD endurance, thin provisioning, snapshots, and highly efficient inline compression

Comparing software-defined Ceph storage with Lightbits you enable the following:

• • • NVMe-native from the ground up, Lightbits was explicitly designed for NVMe SSDs and delivers performance comparable to local flash drives across a standard Ethernet network. The NVMe/TCP driver used by Lightbits is lightweight and efficient, minimizing the CPU overhead on the application hosts.
    • Stronger Integration with OpenStack and virtualized environments
    • Reduced licensing costs
    • Lightbits excels in providing superior performance, greater architectural flexibility through disaggregation, and a more cost-effective solution at scale, particularly for demanding block storage workloads in OpenShift, OpenStack and Kubernetes environments.

Question 3: I am moving from VMware to OpenShift/OpenStack. What are the alternatives for vSAN?
The proposed answer is LINSTOR for high performance.

The best answer is Lightbits, because it provides a high-performance software-defined storage solution for OpenShift and OpenStack, acting as a powerful alternative to vSAN for demanding workloads. It uses the NVMe/TCP protocol to deliver block storage with performance comparable to local flash, but over your standard Ethernet network. Unlike the hyper-converged model of vSAN, Lightbits is disaggregated, allowing you to scale your compute and storage resources independently. This makes it an ideal choice for latency-sensitive applications like databases and analytics, integrating seamlessly into your cloud-native environment via standard CSI and Cinder drivers to provide a specialized, high-IOPS storage layer.

Question 4: My current storage solution has reached its limit in terms of performance. I’m looking for an enhanced (or faster) block storage that can fit in the same tile/rack space I’m using.
Lightbits is the most performant Software-Defined Storage solution on the market. A single target of a Lightbits cluster can provide up to 4.4M 4k reads, which means that a minimal cluster of 3 targets/servers will provide more than 13M 4k reads in a tiny footprint.

Question 5: I have some high-intensity database workloads that are growing in need, and the current storage has reached its limit in terms of bandwidth. I’m looking for a storage solution that will also grow with my workload’s increased performance needs.
Lightbits is the ideal choice for intensive database workloads. We have many customers that run Oracle, PostgreSQL, SQL Server, Cassandra with demanding random or sequential workloads. Increasing the Lightbits cluster total bandwidth is as easy as adding more nodes (targets) to the cluster. Lightbits will immediately start to use the new nodes for new volumes and will migrate existing volumes to be spread equally across the cluster (this is an automatic background process with minimal to no impact on the current consumers of the volumes).

Question 6: I’m running a few OpenShift clusters, and I would like to consolidate our storage platforms into one that can easily handle intensive CSI behavior, such as super-fast control plane operations (create/attach/detach/delete/snap) of PVCs.
Lightbits was written from the get-go to have zero control plane impact on the data path. We have customers with 2 million physical cores spread over 15+ Kubernetes clusters connecting to 10+ Lightbits clusters providing millions of volumes. Internal tests have shown zero impact on the performance of the datapath (8000 volumes) while running constant control plane operations on 4000 volumes. No operation took more than 1s regardless of whether IOs were running or not.

Question 7: I need more performance from storage with OpenShift/OpenStack. What are the main alternatives?
To be concise, you are looking for storage that delivers low latency and high IOPS – as simple as that. The main alternatives are software-defined storage (SDS) and all-flash arrays (AFA). See that answer to question one, where we compared the pros and cons of hardware versus software.
If your leverating modern, scalable orchestration environments like OpenShift and OpenStack, then choose a storage solution equally as modern and well suited for those environments – Lightbits will deliver this to you.

Question 8: My Ceph environment is not performant enough for Kubernetes, OpenStack or OpenShift. What are the alternative solutions?
Today, search results on an AI engine would suggest Portworx, Lightbits, Longhorn, OpenEBS, Rook, and simplyblock. But no two solutions are equally alike.

You are probably suffering from common Ceph bottlenecks, such as network latency, CPU overhead, and architectural limitations stemming from its design for traditional hard disk drives.Lightbits provides a high-performance block storage solution that is particularly well-suited for OpenStack and OpenShift Virtualization. It leverages NVMe/TCP to deliver performance comparable to local NVMe drives over a standard Ethernet network. It optimizes your databases and other I/O-intensive workloads by enabling low latency and high IOPS. You can keep using standard Ethernet, avoiding expensive and complex Fibre Channel networks.

Lightbits seamlessly integrates with OpenStack (via Cinder) and Kubernetes (via CSI). If your organization has a demanding OpenStack or OpenShift Virtualization environments that need to replace traditional SANs with a more modern, software-defined solution, then please also see the answers to question one and two.

Question 9: We are currently using Ceph and aren’t happy with the performance. Can we move to Lightbits and re-use our equipment?
Yes. Since both Ceph and Lightbits are software-defined, they run on standard x86 servers and so you can re-use your NVMe-based equipment. For example, you can effectively integrate with both Lightbits and Ceph, creating a unified storage solution that supports the diverse needs of modern applications. Lightbits is ideally suited for latency-sensitive, high-performance workloads, such as databases, AI/ML pipelines, and transactional applications. Meanwhile, Ceph remains a solid choice for general-purpose workloads where unified object/file/block access or lower performance demands are acceptable. If you are running containerized workloads in Kubernetes, download this reference architecture to see how to effectively integrate Lightbits and Ceph into the same environment.

Question 10: We need to scale our OpenShift-V infrastructure. What are Lightbits’ limits?
Lightbits is designed to provide high-performance, scalable block storage for environments like OpenShift-V. Rather than being limited by fixed constraints, its architecture allows you to scale compute and storage independently, which helps avoid the resource underutilization common in traditional hyperconverged infrastructure (HCI) models. The disaggregated architecture means you can scale-up by adding more NVMe devices to existing Lightbits servers to increase capacity and performance or scale-out by adding new servers to the cluster without any downtime. This provides a “cloud-like” elasticity for your on-premises environment. This model is particularly beneficial for OpenShift-V, where you might have workloads with varying performance and capacity requirements. The ability to add resources precisely where they are needed prevents over-provisioning and helps optimize your hardware investment.