On 20 Jan 2018, at 18:17, Ralph Wetzel theonionbox@gmx.com wrote:
Good morning! I recently discovered, that the latest bandwidth record of my relay (97FC02820912441BC2DFE3ACF433E455714B0AF5) availabe from onionoo.torproject.org for the scale of '1 months' and '3 months' is quite old, from "2018-01-12 10:00:00". I've validated this as well with the data available from the hidden service onionoo servers - same result. metrics.torproject.org seems to display a similar picture - yet it's a bit difficult to estimate from the charts. The node is stated as 'running' (and it was for the whole time) and the weights data is available (low score but score). There's 'noticeable' traffic running through this relay - of same amount as always before. Any idea what's going wrong here? Can I fix this - or is it a network issue?
We deliberately report bandwidth less often in recent Tor versions. We think less detailed bandwidths make it harder to find onion service guards.
We need to fix Onionoo to process the new bandwidths: https://trac.torproject.org/projects/tor/ticket/24155
If you want more detailed graphs, you can run Nyx or any other bandwidth monitor on your relay machine.
T
Hi,
On 20/01/18 10:25, Ralph Wetzel wrote:
As a consequece, I'll consider implementing a recording function into The Onion Box.
When you do this, please make it clear to users that making their fine-grained bandwidth usage information public may harm the anonymity properties of the Tor network.
Thanks, Iain.
Gesendet: Samstag, 20. Januar 2018 um 17:06 Uhr Von: "Iain Learmonth" irl@torproject.org An: tor-relays@lists.torproject.org Betreff: Re: [tor-relays] Onionoo bandwidth recording stopped? Hi,
On 20/01/18 10:25, Ralph Wetzel wrote:
As a consequece, I'll consider implementing a recording function into The Onion Box.
When you do this, please make it clear to users that making their fine-grained bandwidth usage information public may harm the anonymity properties of the Tor network.
On 21 Jan 2018, at 04:03, Ralph Wetzel theonionbox@gmx.com wrote:
Isn't this an inherent contradiction?
No. Making bandwidth information public makes it easier to link onion services with their guards. It might also allow other kinds of attacks.
We think it's safe to release a daily bandwidth figure for each relay. Or, more precisely, relays publish daily bandwidths so we can do bandwidth measurement and statistics. We don't like releasing that level of data, but it would take a lot of development effort to do it differently.
If someone exposes his bandwidth usage information to public access, he already harmed the anonymity of *his* relay.
Relays are not anonymous.
Anonymity is a property of the Tor network, not individual relays. But individual relays can compromise the anonymity of clients that build paths through them, by making it easier for adversaries to find a client using that client's traffic.
Yet, as the bandwidth recording & display is local to the monitoring instance (with no API provided),
I see screenshots of bandwidth on Twitter. And publicly available munin pages on relays. You might be surprised what people release.
even if disclosed to public access, the harm is - according to my understanding of the matter - limited to the node(s) monitored, if at all. How does this (local situation) 'harm the anonymity properties of the [whole] Tor network'?
Relays are not anonymous. Releasing detailed bandwidth can harm the anonymity of clients. See above.
T
tor-relays@lists.torproject.org
-
Iain Learmonth -
Ralph Wetzel -
teor