[Dextrose] [Sugar-devel] Datastore index corruption!

[Bernie Innocenti]

On Fri, 2010-07-30 at 07:11 +0000, Aleksey Lim wrote:
> > Is it in Rarian or in the datastore code itelf? How do we change it to a
> > sane 2-3 seconds delay?
> 
> imo, initiating ds flush (or regualr ds shutdown), and thus setting update
> flag, on power-off button ACPI catch is more appropriate way to handle only
> this usecase (we can all time play w/ timeout and don't find proper value
> that will be good in all cases).

This would help catch a few more cases of data loss, but not all of
them.

So, while it would be useful, I'd complement it with a shorter flush
delay. What's to be gained by delaying writes 60 seconds anyway? How
long does it actually take to flush the index to disk?

-- 
   // Bernie Innocenti - http://codewiz.org/
 \X/  Sugar Labs       - http://sugarlabs.org/