IEEE Computer Society

MSST

Sponsored by the IEEE Mass Storage Systems
Technical Committee
IEEE Computer Society

April 16 — 20, 2012
Asilomar Conference Grounds
Pacific Grove, CA


MSST 2012 Speaker

Dr. Sage Weil, Ceph

Sage Weil
Tutorial Abstract: ceph: distributed storage for cloud infrastructure

Ceph is a fully open source, distributed object store, network block device, and file system designed for reliability, performance, and scalability from terabytes to exabytes. Ceph's architecture is based on an object storage service that provdes a generic, scalable storage platform with support for snapshots and distributed computation. A flexible and well-behaved pseudorandom placement algorithm (CRUSH) facilitates strategic placement to minimize the probability of data loss, intelligent and self-aware storage nodes, many-to-many failure recovery, and proportional and transparent data migration. This storage backend is used (among other things) to provide a simple network block device (RBD) with thin provisioning and snapshots. Native support for RBD in Qemu/KVM, libvirt, and OpenStack make Ceph an attractive storage option for cloud deployments.

This tutorial will include an overview of the Ceph architecture, typical deployment scenarios (including common hardware and software configurations), a detailed discussion of how CRUSH works and how data placement is managed, and a practical discussion of how RBD and radosgw can be used to store and manage virtual machines and object storage in a cloud environment.

MSST Abstract: the ceph distributed storage system

As the size and performance requirements of storage systems have increased, file system designers have looked to new architectures to facilitate system scalability. Ceph is a fully open source distributed object store, network block device, and file system designed for reliability, performance, and scalability from terabytes to exabytes. Ceph utilizes a flexible and well-behaved pseudorandom placement algorithm (CRUSH), active storage nodes and peer-to-peer like protocols to avoid the scalability and reliability problems associated with central lookup tables or gateway servers.

This talk will describe the Ceph object store and distributed metadata architecture, and then focus on the current status and future of the project. This will include a discussion of Ceph's relationship with local filesystems (like btrfs, xfs, and ext4), the file system and RBD clients in the Linux kernel, RBD support for virtual block devices in Qemu/KVM and libvirt, and current engineering challenges.



Co-founder of DreamHost and New Dream Network, Dr. Weil is currently focused on directing the development of Ceph. "Currently directing development of the Ceph open source distributed file system, with the aim of creating a production quality system." He has been prolific in starting new ideas with the most notable being WebRing in the late 1990’s. Dr. Weil has over 18 years in the industry, with a Computer Science Ph.d. UC Santa Cruz, 2007, and Computer Science BS Harvey Mudd, 2000.


Page Updated April 25, 2017