This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "". The branch, master has been updated via 99f063b444f5db8b9e630d3a43488817d654f1d6 (commit) from 40e9abc1303d7cbb7313eac1b0a80d7529b91e92 (commit) Those revisions listed above that are new to this repository have not appeared on any other notification email; so we list those revisions in full, below. - Log ----------------------------------------------------------------- commit 99f063b444f5db8b9e630d3a43488817d654f1d6 Author: Phil Carns <[email protected]> Date: Mon Oct 29 17:48:35 2012 -0400 fill in some text about CRUSH ----------------------------------------------------------------------- Summary of changes: .../simulation-2012-10/simulation-2012-10.txt | 16 ++++++++++++++-- 1 files changed, 14 insertions(+), 2 deletions(-) Diff of changes: diff --git a/code/doc/resilience/simulation-2012-10/simulation-2012-10.txt b/code/doc/resilience/simulation-2012-10/simulation-2012-10.txt index 0801988..0489bb7 100644 --- a/code/doc/resilience/simulation-2012-10/simulation-2012-10.txt +++ b/code/doc/resilience/simulation-2012-10/simulation-2012-10.txt @@ -157,8 +157,16 @@ based algorithms for object placement: These algorithms and their trade offs will be discussed in greater detail in <<sec-rebuild-sim>>. -TODO: somewhere (here or in future work) talk about and reference the Ceph/CRUSH work as -well. +Weil et. al. <<Weil>> developed a related object placement algorithm, +known as CRUSH, that distributes replicated objects over a large +collection of servers. CRUSH is notable for allowing administratorss to +partition devices into hierarchies to reflect expected failure domains +and for allowing administrators to weights to different devices. +Previous studies of the CRUSH algorithms have not focused on the time +required to complete a rebuild, however, but instead focused on the cost of +the placement algorithm itself and the degree to which it minimizes +superfluous data movement following server failure or addition. We will +consider CRUSH algorithms in future work. == Methodology @@ -553,3 +561,7 @@ erasure coding algorithms for quiescent data. - [[[Maymounkov]]] Petar Maymounkov and David Mazieres, “Kademlia: A peer-to-peer information system based on the XOR metric,” in Proc. Workshop on Peer-to-peer Systems, 2002. +- [[[Weil]]] S. A. Weil, S. A. Brandt, E. L. Miller, and C. Maltzahn. + CRUSH: Controlled, scalable, decentralized placement + of replicated data. In Proceedings of the 2006 ACM/IEEE + Conference on Supercomputing (SC ’06), Tampa, FL, Nov. 2006. ACM. hooks/post-receive --