[tor-talk] Why is there a huge gap between Total BW retrieved by Stem, Total shown on Tor Metrics and in http://torstatus.blutmagie.de/?

[Anonymous Kyoto]

Hi all,
① I just try fetching some information in the descriptors consensus from
directory authorities via Stem library. After fetching and sum up the
bandwidth, I got this number: 30841602 kilobytes, which is around 246 Gbit.
It is quite different from the number of around 60GBit/s of bandwidth
history, and 125 Gbit/s for Advertised bandwidth in Tor Metrics:
https://metrics.torproject.org/bandwidth.html.
I know that the bandwidth I fetched via Stem library is the line w in
consensus document stated here:
https://gitweb.torproject.org/torspec.git/tree/dir-spec.txt#n1768

P/S: Code I wrote to fetch bw is:

> from stem.descriptor.remote import DescriptorDownloader
> downloader = DescriptorDownloader()
> print '%s %i ' %(IP, int(desc.bandwidth))


② Next, simultaneously I checked bandwidth of a particular IP (say,
95.215.44.189),
In Consensus, it shows 264 (
http://171.25.193.9:443/tor/status-vote/current/consensus)
In Vote, it shows 604 (
http://171.25.193.9:443/tor/status-vote/current/authority)
and in http://torstatus.blutmagie.de/, it shows 66KB

By reading the Direcroty Protocol, I can figure out the difference between
Consensus and Vote, but Why the last line is also different from both Consensus
and Vote? Do anyone has any idea of how the site gets the 66KB/s number for
bandwidth?

Looking forward to receiving any comment or answer from you!
Yours Sincerely.