Introduction

LeoFS Overview

“LeoFS” is highly scalable, fault-tolerant distributed file system (DFS) for the Web. Different than traditional distributed file system and other DFS — “LeoFS” offers a number of unique benefits to users:

• High Cost Performance Ratio
• High Reliability
• High Scalability

Goals

• LeoFS aims to provide the following advantages:

  • HIGH Cost Performance Ratio

    • Fast - Over 200MB/sec into 10GE
    • A lower cost than other storage
    • Provide easy management and easy operation

  • HIGH Reliability

    • Nine nines - Operating ratio is 99.9999999%

  • High Scalability

    • Build “Huge Cluster” at low cost

Milestones

• DONE - 0.12 (Oct 2012 - Jan 2013)

  • Large Object Support (incl.Streaming/Multi-part/Range requests)

  • Web GUI-Console (LeoTamer - Optional)

    • Cluster manager/monitor
    • Log Analysis/Search

• 0.14 (Feb 2013 - May)

  • Multi-layer Cache (Using SSD)

  • Job Scheduler on the Manager

  • API-related:

    • Increase compatibility S3-APIs#3
    • Support for OpenStack Swift-API

  • QoS System Phase-1 (LeoDenebola - Option)

  • Web-Console (Leo Tamer - Option)

    • Log Analysis/Search

• 1.0 (June 2013 - )

  • Multi-Datacenter Data Replication

  • OpenStack Integration

  • Increase compatibility S3-APIs#4

    • Objects Expiration into the bucket
    • Other bucket operations

  • QoS System Phase-2 (LeoDenebola - Option)

  • Web-Console (Leo Tamer - Option)

    • QoS Integration
    • Log Analysis/Search