Re: Bug#3133: quota files not sparse after running quotacheck
Marek Michalkiewicz:
> Hmm. My copy of the info documentation says that it does, but only
> if the source file contains holes. After quotacheck, it doesn't...
> Is there any reason why cp doesn't always make holes?
The blocks might be reserved with a reason, even if they are currently
all zeroes. For instance, a swap file must not contain holes, because
otherwise the kernel would have to allocate the physical block when swapping
out a page. The allocation might fail, which is not a good idea.
Similar reasons might apply to databases.
Reply to: