[tor-talk] circuit_send_next_onion_skin(): Bug: Unexpectedly high circuit_successes

[Mike Perry]

Thus spake grarpamp (grarpamp at gmail.com):

> ver: branch 0.2.3.x as of jul 17.
> Should I ticket this or append more info?

Ticket and assign it to me. Cc nickm (if you can).

We do need a bit more info in the ticket, though. See below.

> I'm seeing hundreds of these. The current M seems to exceed the m,
> for which m is increased, then M exceeds that... repeat.
>
> [warn] circuit_send_next_onion_skin(): Bug: Unexpectedly high
> circuit_successes (M/m) for guard <nick>.

Hundreds of these on a normal client? Are you doing anything atypical?
Running a hidden service? Running a relay? Running a network scanner?
Using a custom Tor controller?

> There are a few accompanying entries shown below.
> But note primarily that the above style uselessly lists only the <nick>,
> while the below style needlessly lists it.
> 
> [notice] Low circuit success rate M/n for guard <nick>=<fp>.

You get both this message and the above message for the same guard? 

What ranges of values of M, m, and n have you seen?


-- 
Mike Perry
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://lists.torproject.org/pipermail/tor-talk/attachments/20120727/1cc72fd5/attachment.pgp>