GEOM architecture and the (lack of) need for foot-shooting

[Andrey Chernov]

On Thu, Apr 07, 2005 at 10:34:37PM -0700, Marcel Moolenaar wrote:
> Your angle is slightly different from mine. We do share that the on-disk
> and in-core data can differ, but you seem to allow editing of the 
> in-core

Yes I want to allow editing of both, for more flexibility and safety. See 
below.

> data by partitioning tools, while I don't.
> 
> is dropped when the disk disappears. The on-disk data can be modified
> by partitioning tools. The in-core data does not change because of that,
> but the in-core data can be brought in sync with the on-disk data by
> some means (sysctl, ioctl or whatever). The in-core data cannot be 
> edited
> on its own. 

It bring some problems like illegal on-disk modification synced to 
in-core. Since on-disk editing is not controlled (and should not be), it 
may overlap or be incorrect in some other way. But, if you edit in-core 
partition instead, as I suggest, you can do all sorts of checking and 
safety, easily excluding overlaps, etc. I.e. I suggest in-core->on-disk 
sync (which always write checked result) instead of can't be checked 
on-disk->in-core sync.

-- 
http://ache.pp.ru/