[tahoe-dev] pubgrid share size distribution

Brian Warner warner at lothar.com
Wed Feb 23 10:21:40 PST 2011


On 2/23/11 9:54 AM, Greg Troxel wrote:
> 
> In particular, I'm finding lots of shares that fit this find command,
> more than I would expect:
> 
>   find shares -type f -size +10324 -size -10346

Hmm. If you run 'tahoe debug dump-share' against one of those, you
should see the original filesize (before being split up for encoding). I
wonder if there's some stereotypical file type that result in some
unique file size, and somebody's uploading a bazillion files of the same
type.

Also, 'tahoe debug catalog-shares' will print one line per share, and
includes the original filesize in each.. you can probably sort and bin
the output of that to help with this sort of analysis.

cheers,
 -Brian


More information about the tahoe-dev mailing list