[Gluster-devel] [write-beind] why we don't cache if length of data can not be divided by 4096.

Anand Avati avati at zresearch.com
Sat Jan 12 20:16:25 UTC 2008


Li,
 this is legacy code from the time when the inode framework was not around,
and multiple fds and plain calls on filenames could not be related. back
then the aggregate-size was used as a mean to aggregate only broken writes
of fuse. I guess we can remove this piece of code now though it actually
causes no "harm".

avati

2007/12/19, LI Daobing <lidaobing at gmail.com>:
>
> Hello,
>
> a code in write-behind.c, function wb_writev:
> // begin
>   if ((file->size >= conf->aggregate_size) ||
>       (iov_length (vector, count) % 4096)) {
>     wb_sync (wb_frame, file);
>   }
> // end
>
> Why we don't cache(or aggregate) data, if the length of the data can
> not be divided by 4096?
>
> Any reason for this decision?
>
> Thanks,
> LI Daobing
>
>
> _______________________________________________
> Gluster-devel mailing list
> Gluster-devel at nongnu.org
> http://lists.nongnu.org/mailman/listinfo/gluster-devel
>



-- 
If I traveled to the end of the rainbow
As Dame Fortune did intend,
Murphy would be there to tell me
The pot's at the other end.



More information about the Gluster-devel mailing list