Hi, I would like to attend LSF10 to discuss VFS scalability improvements. VFS scalability is about removing global locks from the VFS, dcache_lock, inode_lock, etc. which basically stand in the way of making fs operations to any filesystem and pseudo filesystem "scalable". I do hope to have made some progress towards merging pieces by the time of LSF10, but it won't be complete. I think everybody is agreed that the global VFS locks are going to run out of steam sooner or later (now at 32 threads per socket and no sign of slowing down). Some real users are already hitting problems (eg. see: google's batched iput/dput patches). The questions will just be HOW to improve scalability and how far to go. I have proposed some approaches for improving scalability. These changes need careful reviews from filesystems and VFS maintainers. So I think LSF would be a good place to discuss this. Perhaps even look at specific difficult patches. Thanks, Nick -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
