Good efficiency or good stableness, it's a question. Every cow file system a block update will cause update to the block point to it and recursively to root. Thus these file systems will write more blocks than in-place update file systems(like ext3). This cannot be avoided. But do have some technologies to improve this. NetApp's WAFL uses NVRAM-cache to log write operations. Log-structured file systems always sequentially write to disk to achieve high throughput. BtrFS uses a B+ tree to limit the height of the tree (so can reduce blocks need to write in a COW operation), and uses a log tree to speed up frequently sync operations (I'm not familiar with that). 2009/12/11 Hu Ruihuan <specter118@xxxxxxxxx>: > 2009/12/11 Hu Ruihuan <specter118@xxxxxxxxx>: >> Hi all, >> I am puzzled about a question, everytime when btrfs_writepage is >> called, whethe the noeds in every levels of the fs tree will be >> updated. This is the case as I read in the code, but if this case, >> whether it will give rise to the low efficiency? >> Thanks! >> > refer to the update of nodes, I mean a cow operation > -- > To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in > the body of a message to majordomo@xxxxxxxxxxxxxxx > More majordomo info at http://vger.kernel.org/majordomo-info.html > -- Zhang Jingwang National Research Centre for High Performance Computers Institute of Computing Technology, Chinese Academy of Sciences No. 6, South Kexueyuan Road, Haidian District Beijing, China -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
