- tor-relays - lists.torproject.org

Domain name based policies (was: Call for discussion: turning funding into more exit relays)
by Nicolas Braud-Santoni 01 Aug '12

01 Aug '12

2012/8/1 Roger Dingledine <arma(a)mit.edu>: > On Tue, Jul 31, 2012 at 11:21:01AM +0100, mick wrote: >> Question for tor developers. How hard would it be to change the logic >> (and syntax) of exit policy in tor to allow domain based formulations >> like: >> >> reject *.gmail.com >> reject *aol.com > > Very hard. > > https://trac.torproject.org/projects/tor/wiki/doc/TorFAQ#Exitpoliciesshould… Hi, While I see how allowing wildcards and domains in policies would be more than challenging, wouldn't it be possible to : - resolve domain-names at Tor startup, and get all associated A and AAAA records - Repeat when record's TTL is reached. Of course, it wouldn't work for sites that don't advertise all their IPs. It would also require the Exit node's operator to run some DNS resolver (or trust an external one), but locally running unbound (for example) is quite simple. Moreover, the risk evoked in the FAQ is already present : if I poison an exit node's DNS resolver, wouldn't I be able to replace nytimes.com A record with some bogon, like 0.0.0.0 ? Nicolas

2 1

Re: [tor-relays] Call for discussion: turning funding into more exit relays
by Scott Bennett 01 Aug '12

01 Aug '12

On Fri, 27 Jul 2012 01:23:57 -0400 grarpamp <grarpamp(a)gmail.com> wrote: >>> 4) What exactly do we mean by diversity? >> >> I would look at this almost entirely from a jurisdictional and ISP level. I >> believe the biggest "sudden impact" threats to the tor network are going to >> be from legal changes (jurisdictional, i.e. "save the children, nullroute >> the nodes") and local business policy changes ("sorry tor customers, no more >> tor egress from our DC due to complaints"). > >I'm not sure which thread I mentioned this on so I'll put it here to be sure. >I think one main thing needed is a project to catalog all the current >exits as to their diversity... >Box: ISP/hoster, AS, datacenter, country, upstream AS/Tier-n path, >relay-operator >Relay-operator: country > >Without that, seems like placing nodes amounts to, 'Well, >we don't have any in Iran, let's go there'. If it turns out that >IP is more or less fed as a courtesy from UAE across the >gulf, there's not much gain. Repeat analysis for any of the >above parameters. > >More nodes are probably good, just not all as USA, Equinix, >Level3, with whatever hoster has a rack in all the DC's. I agree completely. But I would also like to add that, aside from Brasil, most of South America is still dark. Central America is not much better either. Many of those states are not especially cooperative with each other, politically speaking, yet they all need the benefits of commerce associated with the Internet. That combination strikes me as beneficial to placement of tor relays in as many of those countries as possible. Much of Africa may be worth closer examination for the same reasons. We really need to keep political diversity in view, especially given the large fractions of the tor network currently concentrated inside a mere handful of politically allied states. The Dictator of the U.S., for example, has already made the threat of shutting down the entire U.S. portion of the Internet, including relaying between other countries, which would certainly have a severely disruptive effect upon tor users all around the globe were it to happen under today's distribution of tor relays. Even more drastic would be if any of, for example, the U.K., France, Germany, or the Netherlands were to follow suit. Having countries like Venezuela, Bolivia, Peru, and Ecuador more tor-populated looks to me like a good thing. Another point I'd like to make is that I don't see why having one 100 MB/s relay is somehow better than having ten 10 MB/s relays or 20 5 MB/s relays. The superhigh-speed relays push operating system limits on the number of connections. Due to tor's design, distributing the workload of such relays across multiple CPU cores is problematic. Olaf Selke got around that problem by running four nodes on a quad-core machine with two IP addresses, but that meant that each node usually ran at less than 15 MB/s. For a superfast setup today, it might mean running multiple 25 MB/s nodes in similar fashion to what Olaf did, rather than a single 100 MB/s node. The benefit to tor users would seem to me to be the same either way, but the multinode method would not satisfy the demand of the funding source, as I understood it. Either way, though, the operating system limits may place keep a lid on the actual tor capacity of a very fast setup. From an infrastructure standpoint, I acknowledge that there can be problems in setting up really fast relays in Latin American countries. However, adding a few relays on the order of 500 - 5,000 KB/s in each Latin American country could probably be done, even if it meant they could only be set up in national capitals, which are mostly megalopolises of several million people, and might involve making special arrangements with the ISPs. Other major cities in some countries may also have the infrastructure to make moderately fast relays possible. Scott Bennett, Comm. ASMELG, CFIAG ********************************************************************** * Internet: bennett at cs.niu.edu * *--------------------------------------------------------------------* * "A well regulated and disciplined militia, is at all times a good * * objection to the introduction of that bane of all free governments * * -- a standing army." * * -- Gov. John Hancock, New York Journal, 28 January 1790 * **********************************************************************

1 0

How to protect yourself from network scanning
by amki 01 Aug '12

01 Aug '12

Hiho, I am hosting a 3-5MB/s tor exit relay but as of today my hoster has closed my server because of network scanning. Is there a known proper way to protect yourself from being used as a network scan relay? I've thought about constructing iptables rules to limit the number of SYN packets for the same host per second or such, but I'm not sure if this is allowed or will get me flagged as a bad exit node. My hoster is quite ok with us generating some abuse complaints per month, but does not want to route network scanning traffic since it is a severe load to their routers. Any help would be appreciated Cheers amki -- "Whatever the reason you're on Mars, I'm glad you're there, and I wish I was with you." - Carl Sagan

5 5

Re: [tor-relays] Call for discussion: turning funding into more exit relays
by Steve Snyder 01 Aug '12

01 Aug '12

Roger Dingledine arma at mit.edu Mon Jul 23 18:58:54 UTC 2012 [snip] >At the same time, much of our performance improvement comes from better >load balancing -- that is, concentrating traffic on the relays that can >handle it better. The result though is a direct tradeoff with relay >diversity: on today's network, clients choose one of the fastest 5 exit >relays around 25-30% of the time, and 80% of their choices come from a >pool of 40-50 relays. [snip] From what I see on the TorStatus pages (torstatus.all.de, blutmagie.de) about a third of the roughly 3000 relays listed are at or below 64KB/sec of demonstrated bandwidth. No doubt some of these are soon-to-be-high-bandwidth servers that are just ramping up, and some are nodes having transitory networking problems. It seems reasonable to assume, though, that most of these low-bandwidth nodes are intentionally low-bandwidth, perhaps on the basis of the Tor doc stating a 20KB/sec minimum. With "80% of their choices come from a pool of 40-50 relays" that leaves a 20% chance for the remaining 2950 nodes. A case for low-bandwidth nodes can be made as a means to dissuade anticipated routing (due to pool size), but it seems from the stats quoted above that there is little chance that 2000+ of these 3000 nodes will ever carry Tor traffic, and thus can be ignored for purposes of traffic analysis. Is there any justification for a low-bandwidth Tor node? And if so, what is the practical minimum bandwidth needed to actually see any traffic?

4 3

Re: [tor-relays] Call for discussion: turning funding into more exit relays
by Roger Dingledine 01 Aug '12

01 Aug '12

On Fri, Jul 27, 2012 at 05:49:34AM -0400, Motoko Kusanagi wrote: > I am very interested in running 100 Mbit (maybe even more) exit nodes at 100$/month, however, a question immediately comes to mind: > When we say "100Mbit exit node", do we imply "really unmetered" traffic at 100 Mbit, or do we mean "semi-reasonable traffic limit at 100Mbit speed". > "Really unlimited", in my experience, oftentimes turns out to be somewhat problematic. I don't think we necessarily need to have the 100mbit all to ourselves (see the other discussions here about shared connections), but I think it needs to be able to push a good fraction of 100mbit throughout the month. So something that can handle 20-30TB/mo is in the right ballpark. Note that many of our current 100mbit nodes are actually gbit nodes that try to limit themselves to a long-term average (95th percentile) of 100mbit. > I could arrange for 50-100 100Mbit exit nodes running 24/7, with a >traffic limit somewhere between 1 and 4 terabytes (will need to scrutinize >my current contract and do some math). Arranging for such a number of >"honestly unlimited traffic at 100Mbit speed" would be...significantly >more problematic. 4TB/mo is good for a 10mbit connection, but not good for a 100mbit connection. You'll end up either needing to hibernate for 80% of the month, or rate limit your relays down a lot. That said, we don't need 50-100 100mbit exit relays from any single group. Just 5-10 would be nice. Maybe that makes the capacity issue easier? --Roger

1 0

Re: [tor-relays] Negative impact of MaxOnionsPending 250?
by Scott Bennett 31 Jul '12

31 Jul '12

On Thu, 26 Jul 2012 17:11:17 +0200 Moritz Bartl <moritz(a)torservers.net> wrote: >We sometimes see "Your computer is too slow to handle this many circuit >creation requests!" on our servers. Scott Bennett suggested to set >MaxOnionsPending to 250 instead of the default of 100, which at least >makes the warnings disappear. [1] > >I am thinking about modifying out torrc template we use as as basis for >all servers. What is the negative impact of setting MaxOnionsPending to >250? Should we instead reduce the bandwidth as suggested in the warning? I don't think so. To the best of my recollection, I've only seen this message in recent versions of tor when it appeared that my node had been chosen for one or more high-volume hidden service operations, where the traffic volume in B/s was still fairly low, yet the number of connections had suddenly climbed to an unusually high level. The data throughput capacity vastly exceeded the actual traffic, but the onionskin-decoding load was sometimes, but probably not continually, very high until the hidden service(s) either shut down or changed their advertised information in the hidden service directory. To lower my node's RelayBandwidth* values would most likely have little impact on the bottleneck. Raising MaxOnionsPending adds a little slack in the queue to handle the occasions where that activity is temporarily backlogged prevents dropped connections and allows operations to proceed normally with only minimal extra delays. Consider a delay of probably no more than a second at the outside and usually less versus the delay of building a new circuit to a hidden service due to a dropped connection. > >[1] https://lists.torproject.org/pipermail/tor-relays/2012-May/001352.html Scott Bennett, Comm. ASMELG, CFIAG ********************************************************************** * Internet: bennett at cs.niu.edu * *--------------------------------------------------------------------* * "A well regulated and disciplined militia, is at all times a good * * objection to the introduction of that bane of all free governments * * -- a standing army." * * -- Gov. John Hancock, New York Journal, 28 January 1790 * **********************************************************************

1 0

Re: [tor-relays] Call for discussion: turning funding into more exit relays
by Zac Lym 30 Jul '12

30 Jul '12

This seems (to me) like an obvious suggestion, so my apologies if it's already been thought up. Why not establish a team/scoreboard system, like those used for distributed computing and BitCoin mining? This elegantly solves a few problems while with minimal resource commitment from the Tor organization. Most importantly, it's a way to pump money into the system without ruining the current atmosphere. Sponsors could easily buy some bandwidth or people can also just donate their own connections and join a particular team. The scoreboard is based on goodwill, not dollars spent. It also eliminates the hassle of setting prices, as teams can compete for dollars and bandwidth provided, essentially setting their own prices. The org could also setup a payment system, like we-pay, that the team admins can configure to deposit funds. It could be set as a proof of work system, paying after the bandwidth has been provided. This also allows a degree of control to prevent abuse from admins trying to juice stats by abusing some network infrastructure, like dummy trial accounts on hosting sites. A group admin could block specific hosts or the Tor project could remove an entire group. Finally, this could allow for the Tor project to create metrics based on things other than speed, akin to how Folding@home scores GPU and CPU contributions differently. Then the project can set anonymity goals for the network (such as location, ISP, backbone provider, etc) and the volunteers will adjust their patterns accordingly. Finally, I would like to strongly suggest taking a cut of all donations to pay for the new infrastructure and Tor research and development. I wouldn't want development to slow on Tor because donations are going to bandwidth instead. Indeed, I would prefer we spend a ton of money on the stenography efforts and usability. Anyway, thanks for the hard work! -Zach Lym

2 1

nickname to fingerprint
by Aichetou Med 28 Jul '12

28 Jul '12

hello tor community my question is : is there an "easy" way to translate relay nickname to fingerprint ? if not is it possible to download descriptors for some specific relays given "Digest" string from consensus file? i mean before launching tor. besides is there a reason for not showing the fingerprint in the consensus. thank you :)

2 1

Script to get top relays, AS and countries weights
by delber 27 Jul '12

27 Jul '12

Hi! Attached is a pretty simple Python script to compute some statistics about relays. Here's the doc: Usage: tor-relays-stats.py <output> [args ...] Where <output> is one of: - countries [FLAGS] relative percentage of the consensus in each countries - as-sets [FLAGS] [COUNTRIES] relative percentage of the consensus in each AS sets - top [COUNT] [FLAGS] [COUNTRIES] top relays according their place in the whole consensus Examples: - To get the top five exit nodes in France: tor-relays-stats.py top 5 Exit fr - To get weights of each AS of all relays in Germany: tor-relays-stats.py as-sets Running de Maybe it has flaws. Maybe it should land in some Git repository. I just felt it might be of interest to other folks. Feel free to comment and hack. -- delber

3 14

Negative impact of MaxOnionsPending 250?
by Moritz Bartl 26 Jul '12

26 Jul '12

Hi, We sometimes see "Your computer is too slow to handle this many circuit creation requests!" on our servers. Scott Bennett suggested to set MaxOnionsPending to 250 instead of the default of 100, which at least makes the warnings disappear. [1] I am thinking about modifying out torrc template we use as as basis for all servers. What is the negative impact of setting MaxOnionsPending to 250? Should we instead reduce the bandwidth as suggested in the warning? [1] https://lists.torproject.org/pipermail/tor-relays/2012-May/001352.html -- Moritz Bartl https://www.torservers.net/

2 1