daiaji
daiaji
@terrytw 那四舍五入不是根本没人用。
> If you have two big A and B files, with a slight difference, bees can still dedupe most of the blocks. It's because bees works at the blocks/extents level....
I understand, it seems that even if I use the NTFS file system on the virtual disk image file, dedupe can hit? Because I am using UEFI, I can't use...
@Zygo Off-topic, does btrfs' copy-on-write copy reduce the performance of continuous file reads and writes or degrade the performance of random files?
@Zygo Another digression, dedupe can reduce the write amplification of ssd? The block being dedupe has never been actually written to ssd? As we all know, the btrfs of the...
> > On SSDs the seeks are replaced with IO transaction overheads, which are almost as expensive as physical head movements on SATA SSD devices. > > Wait, what? This...
@Zygo It sounds that discarding duplicate blocks directly can even reduce the generation of data fragments. After the first dedupe, there will be only non-repeating data blocks on the disk,...
@kakra I know the in-band dedup! But I don't know how that works, but now I know how he works, at least some memory is needed as a staging area.
@Zygo So at least discarding duplicate data is a good way to reduce disk fragmentation, right? Because reflink itself is less than 4k, the bees does not process them, and...