ZFS Storage Alternative
Modern scale-out alternative to ZFS
Simplyblock brings all features loved by the ZFS community, such as copy-on-write, thin provisioning, instant snapshots and clones, and automatic tiering, into a scale-out distributed architecture.
Scale-up Only
ZFS is designed to run on a single storage server. Additional storage devices must be added to the same physical server to scale the available storage pool. While there are workarounds, storage capacity requirements will outgrow the possibility of adding additional storage disks.
iSCSI Only
Remote access to ZFS storage is provided via the iSCSI protocol, which transports SCSI commands over Ethernet networks. While iSCSI “gets the job done,” modern flash storage uses the NVMe protocol as its communication layer.
Tuning is Hard
ZFS has hundreds of tuning parameters to optimize its operation to the requested use case. However, tuning these parameters requires in-depth knowledge of ZFS and the underlying hardware.
Designed for Spinning Rust
ZFS was initially designed for spinning disks and is optimized in its read-write behavior to prevent seeking. Therefore, it writes in large batches every few seconds to conserve IOPS and keep read latencies low. Modern flash devices, however, have millions of IOPS and prefer their queues to be as filled as possible.
What is Simplyblock?
With its NVMe-first scale-out architecture, simplyblock’s intelligent storage platform delivers ultra-low latency, automated management, and remote storage access through the modern NVMe/TCP protocol. Unlike ZFS, which can only scale up on a single server, simplyblock is built upon a scale-out cluster architecture to provide the option of adding cluster nodes whenever necessary. With thin provisioning, copy-on-write, instant snapshots and clones, automatic tiering, and more, simplyblock offers all the features that people love about ZFS while not compromising on IOPS, throughput, and access latency.
Deployment Comparison
Simplyblock provides a unified, scalable, and flexible storage solution to pool and manage flash storage devices, spinning disks, and object storage via the modern NVMe over Fabrics protocols. Simplyblock is highly integrated with Kubernetes through its CSI driver. The scale-out storage cluster can run directly in Kubernetes, virtualized, or on Bare-Metal Kubernetes, and can be scaled up and down without downtime.
In contrast, ZFS uses a scale-up-only architecture, limiting the available storage capacity to the physically available space. When using multiple ZFS storage servers, creating a single, unified view is challenging and requires additional components. Additionally, tuning a ZFS storage server for high throughput and low latency is complex and expensive as outside knowledge is commonly needed.
Architecture Comparison
Simplyblock uses an NVMe-first architecture where data is written to a high-performance NVMe storage pool and then transparently tiered to cheaper storage options such as QLC flash or spinning disks. It provides a unified storage virtualization layer to simplify access and management. The virtual storage devices are accessed via NVMe-oF (NVMe over Fabrics), allowing for low access latency and high throughput.
In contrast, ZFS was designed at a time when spinning disks were the norm. While ZFS can be tuned to work well with NVMe-based flash storage, mixed device setups with flash and spinning disks for high capacity and low latency must compromise on both sides due to the available tuning options. ZFS does not support NVMe-oF and must fall back to the older and slower iSCSI protocol.
Benefits of Simplyblock?
Simplyblock provides infinite storage scalability without sacrificing performance or features.
Simplified Operation
Simplyblock abstracts storage complexities and provides a self-healing distributed storage cluster. Via its CLI, API, or management interface, simplyblock enables self-service provisioning and removes operational overhead.
Cost Efficient Scalability
Simplyblock is optimized for NVMe-based storage devices and cost-efficient hardware clusters. With zero-downtime scalability, simplyblock enables storage capacity growth as needed and dramatically reduces the initial cost of storage hardware.
Scale-Out Architecture
Simplyblock enables infinitely storage capacity scalability with its scale-out architecture. Simplyblock storage clusters can scale to hundreds of nodes and provide the combined IOPS and throughput of all cluster nodes for every virtual disk.
Questions and Answers
How does simplyblock compare to ZFS for Kubernetes storage?
Simplyblock is designed for predictable low-latency block storage, linear scaling, and simpler operations for Kubernetes workloads compared to traditional alternatives.
What changes are required to migrate from ZFS to simplyblock?
Most teams migrate in phases: validate workload behavior, move stateful services gradually, and optimize policies for performance, resilience, and cost.
When is simplyblock a better fit than ZFS?
Simplyblock is typically a strong fit when teams need modern Kubernetes-native storage, better performance efficiency, and easier day-2 operations at scale.
Not sure if simplyblock is right for your team?
Ask AI about migration effort, performance, and total cost of ownership.