Features

A comprehensive overview of every component powering Sarkan’s hyper-converged Kubernetes storage platform.

COMPONENT DESCRIPTION
SarkanSFS Operator High Availability (Raft): Custom (non SaunaFS) RAFT-based election mechanism to manage metadata servers master and shadow roles ensuring seamless failover and preventing brainsplit.

It outperforms the built-in SaunaFS mechanism by completely preventing the "split-brain" problem during major failures. Our solution uses the Raft algorithm based on a strict quorum. This ensures perfect data consistency and a safe failover even in extreme situations—for example, when more than half of the metadata servers go down, or when disconnected nodes rejoin the network with outdated data.

Read docs
SarkanSFS Operator Load and Capacity Optimization: Automated chunkserver rebalancing optimizes data distribution across the cluster. By proportionally allocating chunks across mixed-capacity drives, it ensures all disks fill at the same rate, eliminating bottlenecks and wasted space.
Read docs
SarkanSFS Operator Native SaunaFS Support: Fully supports core SaunaFS features, including advanced replication and seamless configuration of metadata and chunk servers.
Read docs
SarkanSFS Operator Autodiscovery: Instantly integrates storage into the cluster by simply labeling PVs or PVCs. This eliminates manual configuration by automatically turning raw volumes into active components.
Read docs
SarkanSFS Operator Multi-Cluster Isolation: Securely run multiple independent SaunaFS clusters within a single Kubernetes environment, enabling zero-interference "Storage-as-a-Service" for different teams.
Read docs
SarkanSFS Operator External Storage Access: Automatically provisions LoadBalancers and host-port mappings to let external bare-metal servers or VMs consume storage. Additionally, it provides Helm-managed NFS support via Ganesha.
Read docs (Architecture)
Read docs (NFS Ganesha)
SarkanSFS Operator

Build-in monitoring: Integrated metadata and chunk server metric real time prometheus exporters.

Read docs

Collected metrics:

Cluster Space
  • Total space
  • Available space
  • Used / Trash space
  • Reserved space
 Servers & Volumes
  • Metadata servers status
  • Chunkservers status
  • Chunk volumes status
Files and Objects
  • Total files / dirs
  • Trash & Reserved files
 Chunks
  • Total chunks and copies
  • Safe / Endangered / Lost
IO Operations
  • Read/Write count
  • Throughput metrics
 High Availability
  • Server status history
  • Metadata version
  • Raft term status
Compute Resources
  • CPU usage
  • Memory usage
 Hostdisks & CSI
  • Total/Used disk size
  • Disk ready status
  • S.M.A.R.T. health
SarkanSFS Operator Monitoring dashboard: Includes a ready-to-use Grafana dashboard for enterprise monitoring. Track I/O throughput, latency, and cluster health in real-time through a professional "single pane of glass" view.
Read docs
SarkanSFS CSI File System Support: Provides native support for the SaunaFS file system, enabling dynamic volume provisioning and seamless integration within Kubernetes.
Read docs
SarkanSFS CSI Multiple Access Modes: Supports all standard Kubernetes access modes (ReadWriteOnce, ReadOnlyMany, ReadWriteOncePod), including full ReadWriteMany (RWX) support for simultaneous read-write volume mounts across multiple nodes.
Read docs
SarkanSFS CSI Volume Lifecycle Management: Fully supports dynamic volume provisioning, cloning, and seamless capacity expansion.
Read docs
SarkanSFS CSI Snapshots and Restore: Supports volume snapshot creation and fast restoration for reliable data backups and recovery.
Read docs
SarkanSFS CSI Mount Corruption Detection & Recovery: Automatically detects broken mounts and remounts volumes for newly scheduled pods. Existing pods require an application-level restart to regain access to the directory.
Read docs
SarkanSFS CSI Block Volume Support: Dedicated support for raw block volumes, optimized for gateway use cases.
Read docs
SarkanSFS CSI Multiple File Systems Support: Offers compatibility with various file systems. In addition to native SaunaFS, it fully supports ext3, ext4, and xfs when provisioned in block volume mode.
Read docs
SarkanHD CSI
  • Physical Drive Management: Automatically detects, formats, and exposes raw server drives directly as Kubernetes volumes.
  • PV/PVC Management: Dynamically creates and manages PersistentVolumeClaims, ensuring proper binding to corresponding Persistent Volumes.
  • S.M.A.R.T. Monitoring: Continuously tracks and reports the hardware health of physical storage media using S.M.A.R.T. diagnostics.
Read docs
SarkanSFS Go API Client
  • Automation & Integration: Enables direct native communication with the SaunaFS cluster. It fully automates file tree management, supporting standard and recursive directory operations, including the creation of nested paths.
  • Metrics Reporting: Delivers precise, real-time cluster health and filesystem metrics. It directly queries the server for critical data, including available space, trash size, and total file and node counts.
Sarkan Kubernetes Playbook
  • Kubernetes Cluster Automation: Automates the complete deployment of Kubernetes environments using Kubeadm, including seamless setup of the Containerd runtime.
  • Networking & Load Balancing: Ensures high availability with control-plane traffic distribution and integrates MetalLB to provide native LoadBalancer functionality (external IPs) for on-premise clusters.
  • Monitoring Stack: Automatically deploys a comprehensive, ready-to-use cluster monitoring and visualization system powered by Prometheus and Grafana.
Sarkan Management Pack Provides native integration with VMware Aria Operations for automatic discovery, metric collection, and clear visualization of Sarkan environment components.
Read docs
Sarkan Content Pack Centralizes and analyzes cluster logs within VMware Aria, offering pre-built dashboards, queries, and alerts to significantly accelerate troubleshooting.
Automated OVA Deployment Centralizes and analyzes cluster logs within VMware Aria, offering pre-built dashboards, queries, and alerts to significantly accelerate troubleshooting.
Community Support Community Support: Driven by the community with best-effort, low-priority response times. All community tickets are synced one-way to the Enterprise tracking system.
Enterprise Support 24/7 Premium Support: High-priority issue resolution with round-the-clock 24/7 coverage. Includes direct access to dedicated email and phone support channels.