Hi Sage. On Wed, May 14, 2008 at 06:35:19AM -0700, Sage Weil (sage@newdream.net) wrote:Well, it depends... If we are talking about single node perfromance, then any protocol, which requries to wait for authorization (or any approach, which waits for acknowledge just after data was sent) is slow. If we are talking about agregate parallel perfromance, then its basic protocol with 2 messages is (probably) optimal, but still I'm not convinced, that 2 messages case is a good choise, I want one :) I.e. having more than single node to be failed? Google uses 3-way replication, but I can not see any factor, which will force people from lowering failure recovering expectations. -- Evgeniy Polyakov --
| Klaus S. Madsen | Regression in 2.6.25-rc3: s2ram segfaults before suspending |
| Dan Hecht | Re: + stupid-hack-to-make-mainline-build.patch added to -mm tree |
| Balbir Singh | Re: 2.6.23-rc7-mm1 - 'touch' command causes Oops. |
| Arjan van de Ven | [patch] Add basic sanity checks to the syscall execution patch |
git: | |
| Nicolas Pitre | Re: [PATCH] diff-delta: produce optimal pack data |
| Catalin Marinas | Re: hgmq vs. StGIT |
| Mark Levedahl | [PATCH] git-clone - Set remotes.origin config variable |
| Junio C Hamano | Re: tracking repository |
| Richard Daemon | Re: booting openbsd on eee without cd-rom |
| Matt | Setting up a virtual hosting machine w. SSH/SFTP accounts - pitfalls/experiences? |
| Paul Greidanus | [Fwd: Open-Hardware] |
| GVG GVG | ssh_exchange_identification: Connection closed by remote host |
| Jim Winstead Jr. | Re: Root Disk/Book Disk Compatibility |
| Paul Douglas Page | Where is mkfs? |
| Howard Wei-Hao Pan | [Q] Does Linux work with PCMCIA devices? |
| Timothy L. Kay | Use PERL rather than C for system commands? |
