Re: TAKE - fix bug 193, oops in fsstress on extended attributes

[George Georgalis <georgw@xxxxxxxxx>]

On Mon, Dec 02, 2002 at 02:31:38PM -0600, Steve Lord wrote:
>when logging attribute extents, use the correct size for the allocation!
>This got introduced back on Oct 10th when in memory extents where
>changed to be host byte ordered. An incorrect size was used on an allocation.

Just curious, what sort of bugs might surface on a heavily stressed
disk that would not otherwise come up? I know this is a very general
question, sorry, I'm only looking for comments; nothing comes to mind,
short of a 'data is faster than the disk' scenario, but even then wouldn't
the problem be handed to memory/virtual memory?

// George


-- 
GEORGE GEORGALIS, System Admin/Architect    cell: 347-451-8229 
Security Services, Web, Mail,            mailto:george@xxxxxxxxx 
Multimedia, DB, DNS and Metrics.       http://www.galis.org/george