[tor-relays] new VPS bridge bandwidth under-reported

starlight.2015q1 at binnacle.cx starlight.2015q1 at binnacle.cx
Mon Jan 5 09:40:47 UTC 2015


Just setup a new bridge running
and it's working fine.

The bridge is running in a Linux container
VPS and appears to have an iptables
traffic-shaped bandwidth limit of 400KB.
Can browse and download files through
it with decent performance using obfs4.

However self-measurement of bandwidth,
after starting at around 200Kbytes
has steadily declined until now its
showing 8KB and has lost the "fast"
flag.  At this point the bridge
has yet to attract any traffic other
then my testing usage--but it works
just fine nevertheless.

I suspect that latency for small transfers
is fairly bad (in the sense of milliseconds,
rather than seconds) and that this may
be distorting bandwidth metrics.  Also
the number of bytes transferred is
close to zero, which does not seem
likely to help the situation.

Should I be concerned about it?  Anything
anyone can recommend to correct it aside
from looking another VPS provider?
Does the advertised bandwidth have
much effect on whether the bridge will
be disseminated by the bridge database
system?  No "stable" flag yet, but not
enough time has passed for this to


More information about the tor-relays mailing list