On Tue, Jun 27, 2023 at 11:12:01AM -0500, Eric Sandeen wrote: > Ok, but what is the reason for zeroing out the blocks prior to them being > written with real data? I'm wondering what the core requirement here is for > the zeroing, either via fallocate (which btw posix_fallocate does not > guarantee) or pwrites of zeros. Note that even a plain truncate will zero the data visible to the user. I could see this tring to reduce fragmentation by making sure the whole file extension is allocated together insted of split up as the difference processes write their areas. But is there data showing this fragmentation even happens and actually hurts?
