- tor-relays - lists.torproject.org

Google Compute Engine rejected as relay?
by Greg 22 Aug '15

22 Aug '15

Hi, I tried to spin up a relay on GCE a few days ago, and I found that it was outright rejected with a message like "Authdir is rejecting routers in this range". I don't have the IP handy now, but I could easily get another ephemeral IP. I thought I came across a thread saying that there was an attack on the tor network originating from GCE, and that's why it got blacklisted. I'm not finding that thread now. But is GCE going to be removed from the blacklist? I realize it's not a very economical place to run a relay. -Greg

5 6

Re: [tor-relays] Calling for more Exit Relays
by spiros_spiros＠freemail.gr 21 Aug '15

21 Aug '15

Στις 21 του Αυγούστου 2015, 13:21, Sharif Olorin <sio(a)tesser.org> έγραψε: >Could you estimate the number of abuse complaints you receive, or the >amount of time you need to spend responding to them - and how many >exits for how long, for context? I am the operator of exit relays in Iceland, England and Czech republic (as well as some non-exit relays where the providers are less friendly to Tor/don’t understand Tor). The nodes have been built in the last six months (financial circumstance only permitted me to start contributing recently), and I hope to run them for many years to come. As such they are all Debian Linux based with administration via Chef, so I probably spend an hour per month on the admin for all of them in total. Abuse complaints generally come in the form of automated email from Webiron and similar services, via the hosting providers. In one case, the hosting provider is happy to change the PTR record and abuse email in the WHOIS to an address that I control, so that they don’t bother forwarding the mail. I get about 5 of these per month for all of the exits, and no action is necessary. In rare cases, the hosting company based in the Czech republic asks me what steps I have taken in response to the abuse report. In this case I respond quickly with a polite message with the actions I’ve taken (blocking the host in the firewall, temporarily restricting the port in question or agreeing with the complainant that they will block the host). I spend about an hour per month on this task as well. The most frustrating part of this is when one of the exit nodes attacks a honeypot or connects to a sinkholed DNS domain, and I have to query the NIDS to make sure it is legit Tor traffic and my boxe s aren’t compromised. The other category of abuse is far more rare but takes a lot of time, and that is legal threat from a company or state entity. I’ve been very lucky and so far only had two such cases, both of which required multiple emails to be sent and staff to be educated in the purpose and functionality of Tor to make the issue go away. These cases took a few hours of work and worrying each. All of my exits have a web page on port 80 explaining that they are a Tor exit and what that means. Hope this is helpful. Spiros

4 4

Are BWauths punishing 'bad' relays? (not publishing measurements)
by nusenu 18 Aug '15

18 Aug '15

Hi, this is actually a question to BWauth ops. even though the new onionoo measured flag data is not* yet completely rolled out (since it has to be deployed for > a week to assign the flag to all relays), I would have a question since the data says that all relays in the previously detected AWS family [1] are unmeasured. I'm wondering whether this is a pure coincidence or on purpose (by BWauths) or by some other yet to be found reason? thanks! [1] https://lists.torproject.org/pipermail/tor-talk/2015-July/038623.html The "spikes" in the following table brought me two that group again. The table shows unmeasured running relays by the month they joined the network (note this is based on preliminary onionoo data) The "spikes" are caused by relays mentioned in [1]. +-----------+---------+ | joinmonth | #relays | +-----------+---------+ | 2015-07 | 3 | | 2015-06 | 4 | | 2015-05 | 2 | | 2015-04 | 12 | <<< AWS family part II | 2015-03 | 4 | | 2015-02 | 35 | <<< AWS family part I | 2015-01 | 2 | | 2014-12 | 2 | | 2014-11 | 8 | | 2014-10 | 1 | | 2014-09 | 3 | | 2014-08 | 3 | | 2014-07 | 2 | | 2014-06 | 2 | | 2014-05 | 1 | | 2014-04 | 4 | | 2014-03 | 1 | | 2014-01 | 2 | | 2013-11 | 1 | | 2013-02 | 1 | | 2012-04 | 1 | | 2012-02 | 1 | +-----------+---------+ onionoo data from 2015-08-18 18:00:00 (thecthulhu instance)

2 2

No guard flag?
by Tor Tor 18 Aug '15

18 Aug '15

Hi, I'm running a relay node [1], and it's been up for a few days over 68 days. I was thinking it would get the guard flag around then. It has yet to get the flag or have non-zero guard probability. Any idea why that is? Or what I could do to get it to be a guard? 1) https://globe.torproject.org/#/relay/741488D89B860E59D6391ACA27A157E87EB533… Thanks, G

5 5

Guidelines for lifetime of a bridge?
by Tim Sammut 17 Aug '15

17 Aug '15

Hi. Is there a guideline for how long a bridge should exist on a particular IP address? For example, does it make sense to keep a bridge on one IP address forever? Or is it better to move a bridge to a new IP address periodically, perhaps every 120 days? I ask because I saw traffic to my bridge ramp-up fairly steadily, and then quickly drop-off to a low number of clients per day. Thanks! hope you are all well t -- Tim Sammut ~ @t1msammut ~ tim(a)teamsammut.com Ford-Mozilla Fellow at Amnesty International

5 5

urras? [was: tor network "loses" ~50 relays/day due to bw auth problem]
by coderman 17 Aug '15

17 Aug '15

On 5/31/15, Jannis Wiese <mail(a)janniswiese.com> wrote: > ... > At the moment I just see urras missing from the consensus... i would like to report a missing Tor-DA to the authorities. ;) best regards,

2 1

do not run Tor client and OR relay together!
by starlight.2015q3＠binnacle.cx 16 Aug '15

16 Aug '15

Anyone who has configured a Tor SOCKS5 client to run in a 'tor' instance that also operates as an OR relay should isolate the client to a separate client-only process. The client function disturbs relay traffic forwarding in a manner that lends itself to confirmation analysis. See bug 16585, particularly comment 5 and onward: https://trac.torproject.org/projects/tor/ticket/16585#comment:5 Perhaps read comment 10 first.

7 8

Multicore, bandwidth, relays, capacity, location
by grarpamp 13 Aug '15

13 Aug '15

On Wed, Aug 12, 2015 at 9:16 AM, Thomas White <thomaswhite(a)riseup.net> wrote: > For relays, being able to make more use of available bandwidth would > vastly increase the network speed, furthermore make home clients see > an improvement in their daily Tor usage. It also benefits hidden > service people as they can then run their Tor processes on more than > one core and thus handle higher volumes of traffic. Tor appears maybe operating at 50% of bandwidth capacity... https://metrics.torproject.org/bandwidth-flags.html https://metrics.torproject.org/bandwidth.html https://metrics.torproject.org/bwhist-flags.html If that's true, more bandwidth won't have any end user affect. So for the moment it might be worthwhile to study reducing node based latency by ensuring relay CPU's have similar free capacity available for circuit negotiation, interrupt processing, and other tor functions. And to consider not dropping more bandwidth weight into already saturated jurisdictions, but dropping some into the Southern Arc and Eastern Zones in an effort to avoid Orion's Belt... https://metrics.torproject.org/uncharted-data-flow.html [Yes I broke the thread because "no subject" and gmail are even lamer than I].

2 1

IPv6 vs IPv4 exit policies
by spriver 12 Aug '15

12 Aug '15

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hi, I just activated IPv6 support for my two exit relays today, but I do not unterstand/misconfigured the exit policies. I just want to open certain ports at IPv4 (the common known reduced exit policy) and open all Ports at IPv6 except 25. How do I configure such a thing? Current sample config is: [snip] IPv6Exit 1 ExitPolicy accept6 *:* ExitPolicy reject6 *:25 [full reduced exitpolicies snipped out] ExitPolicy accept *:50002 # Electrum Bitcoin SSL ExitPolicy accept *:64738 # Mumble ExitPolicy reject *:* But at Globe only this is visible: https://globe.torproject.org/#/relay/F5B1FC9038A5A65FF16D6729AAB2AEDD67F D2F2A https://globe.torproject.org/#/relay/D9D7A9C203C99945D0DCBD545A20C0CB936 7C742 Can someone help me out there? Cheers! spriver -----BEGIN PGP SIGNATURE----- iQIcBAEBCgAGBQJVyNfsAAoJEMkUf8VoejgPiEEP/jS0fkncI8EWdnNm6mWvOB2Z GybS2ZpzKMYzXTx277gp5AXsySBYsMjVDpxKLsWnfhejbKiv5SxgOYipnfROLoyd uTiqP+Ha+SB83DLM9/x9/OWOImT5qdEhKHVZVrW5UeQQJ9kZPO2R7eQ5w2aQfpVd wJ5nv8iSOuL79LPDVidNcAbFrqZ07R1DW/XUMXdlgdWjNd3ToJEBxussIQFMkQLH l6mQf8lF44Hau+udwiOvmKW6xjkMMPcg16jd/hHmimcQ1pgYV0sVS3XKxJifjKKQ VLIGa1oR1rV6rJEpE+F+6f+8HGJsoilda+eDMdgLzkD0Mi+/kHNbKEokyhAEqmxd fTyY5tt10ypXprYr75F6KEvBEP9h0qHY1NRt38YbLW2J4UCGDcW9ZB3xggFCa2iz 14twdT7qnyu8QuFUzox4DaxEyzeTeZXkKpjHsNABAPlGgenDhDkNldeuQnv4n3xc mBGhoobtDKRsi9aKg+9n1qgePOk7kJVUzEyi1Vwk8RdXwu8AyucejVoC7m7cgakW bFWGyNumezGlY2jMz+ARAEDJfG0LrgpEalJ06qZovUibLBCcOludI0Xdp7Y04vH0 PnWotybn7YZjOa69oJ7HGxiAHEoNgT4SeRM/NIvVWTOdgLPbE1Nr2o1drYpmMyjR aoSklTseBiKF+b1bex1L =Bo5q -----END PGP SIGNATURE-----

3 2

Re: [tor-relays] Guard flag flapping
by torrry＠Safe-mail.net 10 Aug '15

10 Aug '15

> So we now have the bandwidth, IP, and dirport of the fastest exits. With this list in hand, I just needed to form a proper URL, wget each one, and grep out the transfer speed: > > http://37.130.227.133:80/tor/server/all 1.17 MB/s > http://176.126.252.11:443/tor/server/all 4.54 MB/s > http://176.126.252.12:21/tor/server/all 666 KB/s > http://77.247.181.164:80/tor/server/all 111 KB/s > http://77.247.181.166:80/tor/server/all 330 KB/s > http://195.154.56.44:80/tor/server/all 3.65 MB/s > http://77.109.141.138:80/tor/server/all 2.20 MB/s > http://96.44.189.100:80/tor/server/all 13.4 MB/s > http://197.231.221.211:1080/tor/server/all 347 KB/s > http://89.234.157.254:80/tor/server/all 295 KB/s > > I'm not seeing anything immediately, although I need to run it on a larger set. There's no smoking gun so far though. Some of the speeds are a bit slow, but nothing low enough to explain the extremely low measured bandwidth these relays are getting. The current BW auth measurement results are around 1.0MBit/s for greendream848. I had a couple of measurements in the 300-500KBit/s range. So, if the auths heavily weight towards low individual measurements, things might make sense. Maybe one of the BW auth guys can comment on how the total measurement result is cooked up!?

2 1

Re: [tor-relays] Guard flag flapping
by torrry＠Safe-mail.net 10 Aug '15

10 Aug '15

> Thanks for running the tests. Which exit nodes led to poor performance? I would like to try to reproduce any performance problems. I did not record the nodes (they were in Europe). A simple test you could run on your server is fetching directory info from nodes that have directory functionality enabled. wget http://<relay IP>:<dir port>/tor/server/all e.g.: wget http://176.126.252.11:443/tor/server/all You can get a bandwidth-sorted list of nodes at: https://torstatus.blutmagie.de/ There is a column that has the directory port. > How would you measure performance between my node and a given exit without being influenced by the properties of the middle relay? You can only set me as an entrynode, and you can't pick a specific middle, You are wrong. :-) You can build arbitrary circuits by hand. There are libraries like Stem, Txtorcon, TorCtl and there is a text based, SMTP-like protocol that you can use directly: https://gitweb.torproject.org/torspec.git/tree/control-spec.txt > so how would you know that the low performance was my node and not the random middle relay? Your node would be a non-random middle relay. > > The bandwidth auths probably downrate the measurement results of your server severely because of those slow connections. > > Probably? How can we investigate further? AFAIK, the raw bandwidth auth measurements are not published, only the total result.

4 3

Re: [tor-relays] I.P. being Blocked?
by Carlin Bingham 10 Aug '15

10 Aug '15

On Mon, 10 Aug 2015, at 05:19 AM, Roman Mamedov wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > On Sun, 9 Aug 2015 13:02:14 -0400 > Zack Weinberg <zackw(a)cmu.edu> wrote: > > > several "this IP is a source of spam" blacklists indiscriminately list _all_ > > Tor relays, whether or not they are exit nodes. > > Now this is just unnecessarily FUDish, at http://bgp.he.net/ one can check > their IPs against 49 RBLs, I checked several of my long-running relays' IPs, > and they are on none of the 49. > Try MX Toolbox's blacklist check[0], it searches over 90 blacklists and you'll find at the very least you're on one of the Tor-specific lists. [0] https://mxtoolbox.com/blacklists.aspx -- Carlin

5 4

I.P. being Blocked?
by Kurt Besig 09 Aug '15

09 Aug '15

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Without going into any detail at this point I'm wondering if others, running a simple middle relay, have encountered any governmental harassment as a direct result of running a relay. The blocking of state agencies e.g, CA.SDI.gov, CA.DMV.gov? Over the past month I've noticed regardless of which browser, machine on my LAN, and wired or wireless connection utilized it's become impossible to connect to any ca.gov site. I've researched my router, cable modem, and firewall settings, all appear to be fine. Spoke with my ISP, for whatever worth that might have, and was assured no blocking or filters were in place that would keep me from reaching these services and my settings looked fine from what they could see from their end. I'm able to connect to ca.gov sites using my phones 4G network or through Tor, however for some reason my home network and associated I.P. is poison. Any experiences/input would be appreciated. Thanks -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (MingW32) iQEcBAEBAgAGBQJVx4UrAAoJEJQqkaGlFNDPILMIALq/zYHqdm/QpThUI/vbr74L Q5lzLiVJBbxOxC3nWX/fuJL35FyS2FwOY4K+piI0EcDpY+wZ/m2hZh7wJLzwh9ok z0j0Dcp7GL3N2c64UP/7XFCrlaNtMm/0ke1xsxUkd0NahZNDCqv7BOkR9IX5yAfx lK5qOw9fpjyzVIdLd1LM1VArWGoEI6dR8PI3rmmGmzP6m9OGC8MxqW2rx+g1p0C5 0F+wDgnmM9rUa2VcgGXf7F4lJkEpUOtjwrsrasclLE8WbwBVv47a84eSSCLKCnUm rD5fUjFBgcaJtrVRgrwydWBcCuIujfxyiCz+Ku8+jtazB5cMdeiZjZN1pgRgKS8= =z1cc -----END PGP SIGNATURE----- --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus

3 4

Re: [tor-relays] Guard flag flapping
by torrry＠Safe-mail.net 09 Aug '15

09 Aug '15

-------- Original Message -------- From: Green Dream <greendream848(a)gmail.com> Apparently from: tor-relays-bounces(a)lists.torproject.org To: tor-relays(a)lists.torproject.org Subject: Re: [tor-relays] Guard flag flapping Date: Fri, 7 Aug 2015 21:49:16 -0700 > I've learned from this thread that the Guard flag flapping is a direct result of the low measured bandwidth. I still have no idea why the measured bandwidth is so (terribly, comically) low. The fact that it's so wrong is somewhat telling. Actual performance on the network is like 1000 times better than what the measured bandwidth says! Something feels very broken here. I ran some tests against your node. While performance is generally very good, it has very low performance connecting to some exit nodes. The problem is likely that your ISP is routing some traffic via an overloaded peering point. Some ISPs have been know to do that on purpose, e.g. to make connections to Netflix or Youtube extremely slow while claiming not to do throttling. The bandwidth auths probably downrate the measurement results of your server severely because of those slow connections.

2 1

BoingBoing: What happened when we got subpoenaed over our Tor exit node
by starlight.2015q3＠binnacle.cx 08 Aug '15

08 Aug '15

http://boingboing.net/2015/08/04/what-happened-when-the-fbi-sub.html check out comment #2 as well as the blog article

4 3

Re: [tor-relays] Guard flag flapping
by torrry＠Safe-mail.net 08 Aug '15

08 Aug '15

-------- Original Message -------- From: starlight.2015q3(a)binnacle.cx Apparently from: tor-relays-bounces(a)lists.torproject.org To: tor-relays(a)lists.torproject.org Subject: [tor-relays] Guard flag flapping Date: Sat, 08 Aug 2015 14:39:48 -0400 > My advice is that this QWest service is third-rate > and rather than bleeding for the length of the > contract, run for the hills! > > If you are still in the 1-month cancellation > period (relays are about a month old), terminate > the service immediately and place an order for > Verizon FiOS. Verizon is one of those ISPs who have played routing games, slowing down Netflix and other traffic that happens to use use the same routing: http://www.extremetech.com/computing/186576-verizon-caught-throttling-netfl…

1 0

Guard flag flapping
by starlight.2015q3＠binnacle.cx 08 Aug '15

08 Aug '15

[apologies to all for thread-breaking, am really going on hiatus but the horror-show performance GBE topic was too darn interesting--last post I promise!] >I would call it a dedicated gigabit link. This is >probably up for debate. The provider's overall >capacity is very likely not [number of customers] >x [1 Gb/s] but I've never witnessed signs of >throttling or over-subscription. Pulling a whois on the IP 184.100.166.110 and a quick Google turned up that you are probably a bleeding-edge subscriber to this new Quest service: https://www.centurylink.com/fiber/ Running a filter on Blutmagie hostname contains quest shows that your relays are, by a huge margin, the fastest of about a dozen. My advice is that this QWest service is third-rate and rather than bleeding for the length of the contract, run for the hills! If you are still in the 1-month cancellation period (relays are about a month old), terminate the service immediately and place an order for Verizon FiOS. FiOS costs more, but sells real bandwidth instead of imaginary bandwidth. You can run a relay on a 75/75 MBit link (like mine) and obtain respectable bandwidth ranking. FiOS goes up to 1/2 GB for about $300-400 per month but the bandwidth is real and the network is good-to-excellent. Either that or switch to leasing a colo server in Germany or some other country where bandwidth is extremely cheap, or perhaps finding a cheap bandwidth colo in the US. QWest GBE looks like a turkey.

1 0

Guard flag flapping
by starlight.2015q3＠binnacle.cx 08 Aug '15

08 Aug '15

>The problem is likely that your ISP is routing >some traffic via an overloaded peering point. Running traceroute -w 1 -q 5 -A <bwauth-IP> 1400 may provide more detail on this issue. The -A option causes 'traceroute' to show Autonomous System numbers with each route hop, where ASNs are Internet routing domains associated with different network providers. One can readily see where big increases in latency are occurring along the path and what peerings are associated with it. Note that some big increases are justified--such as with a hop over a transatlantic cable. Setting -q 5 (or bigger) causes traceroute to attempt each hop that many times. If you see asterisk characters, that's packet loss and may indicate congestion points. Setting the packet size to 1400 will increase the probability that network congestion will impact the results. Try pulling up https: // torstatus dot blutmagie dot de which shows IPs together with other relay attributes and experiment tracing to various major relays.

1 0

Guard flag flapping
by starlight.2015q3＠binnacle.cx 08 Aug '15

08 Aug '15

>. . .have physical access to the . . . ONT. > Doubt this is the case, but on the off chance that the ONT is configured to flow IP traffic over a coaxial cable attached to a CPE router (customer premise equipment), the configuration should be changed to deliver data over a Gigabit Ethernet directly to a good-quality GBE switch provided by you. New eight-port Netgears are surprisingly good managed switches for around $100. For under $600 you can purchase used 24-or-48-port Nortel/Avaya 5510/5520 switches on eBay. Though old, these enterprise grade switches have astounding performance characteristics due to the Broadcom silicon switch-fabric chipset. Were around $7000 new when they were current-gen. All above switches have VLAN support so one switch can operate as multiple isolated networks. I run a Verizon FiOS 75/75 and the default setup is for Internet traffic to pass to a Verizon supplied Actiontec router over a coax cable. Fortunately I hung out with the installer and found out up front--had it switched to the GBE from the outset. For any VZ customers reading this, note that you can connect the Actiontec to your network so it can obtain TV programming information. The coax will continue to carry video and phone data which are on separate fiber optic wavelengths/channels.

1 0

onionoo.thecthulhu.com: 500 Server Error
by nusenu 08 Aug '15

08 Aug '15

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hi Thomas, in case this is not known to you yet: There appears to be a problem with your onionoo instance: https://onionoo.thecthulhu.com/summary?limit=1 Error 500 Server Error -----BEGIN PGP SIGNATURE----- iQIcBAEBCgAGBQJVxZ6AAAoJEFv7XvVCELh05cUQALASqTfj6gZTb8gd73WT/9mB gcUHqhl6scGBNW9Vl88X0uAk55HdO36nqOc/rTv9GaMSFQQ1bgO1nUIFWTvDcpNG /yLTHwWOshoaVNLmL6kJV4KiaLDtUjA3jY8knh59mIYBjL6gLaIKjp8c55i4HUzM UmYa9vPVVOmdWhAuIoIImbEfQqRcR61Ig0NBGheYWF7AAPBgQUv9c+c4fCYOBnla hs9pEMQJ/5u860RJqfETbEe/JVwD3zD+gJL9S7tshQNWKFAShPElgVMRvPu0aRK2 7IIG4szdinqXKd/7Ii9ZtllhWGu9vDGYeFP+9U/sHg4RGZ7bJhLwsMd5VTOgLGpb b5IEPxfTxfSN27FJig+bj6NNCtsVEmYKFayeWhAbml2Rip60a9fc+OBd9tGP6jlb pGCtox02kVUt+7zFJv1sqdf3Qp05RwvETr0it5d0V2+Lm68MRfVZDZ7O24IOTaYp eG6gBMnkmL2hNbgNgUMlvqOOiLYOyjlzhJx1toFLJcdWm9yfSTn90DrfaDFzMacF RRa7oW/mCzo0vseYhTWKJO8w4HJZJUbumZ0w11liqJbJQc1bg4nSVHurVf88Fb6H 5E80IsSFqNA1/vp15tTQsuAkTCK/ycLkKpPYd4Sz8ZEGiKXVqCNszG0QX8MCgk9Q QHz0wwggcg9wGiHFG13l =9H2Z -----END PGP SIGNATURE-----

2 2

Guard flag flapping
by starlight.2015q3＠binnacle.cx 08 Aug '15

08 Aug '15

>I had already run tests with both speedtest-cli >and iperf3. This server consistently achieves 200 >to 300 Mb/s in both directions, with both relays >still running, and on some runs is hitting over >800 Mb/s. Final caveats: If the server is on a shared gigabit link, performance may not improve. The difference between a dedicated gigabit and a shared gigabit is immense. One expects an ISP to load a shared GB links as heavily as possible--I.E. up to the point where customers complain and terminate service. True dedicated gigabit links are expensive. If it's running in a VPS that's showing a gigabit link speed, the number is fiction.

2 1

Guard flag flapping
by starlight.2015q3＠binnacle.cx 08 Aug '15

08 Aug '15

One more errata: The sense of this is negative, so the current setting you have is correct: net.ipv4.tcp_no_metrics_save = 0 The double-negative got me.

1 0

Guard flag flapping
by starlight.2015q3＠binnacle.cx 08 Aug '15

08 Aug '15

comments backward, but the sysctls are correct * 25 milliseconds or 3125000 (for write) * 40 milliseconds or 5000000 (for read)

1 0

Guard flag flapping
by starlight.2015q3＠binnacle.cx 08 Aug '15

08 Aug '15

Ah, forgot to calculate the default TCP windows for you link speed rather than mine. So that's 125000000 bytes/sec ( 1 gigabit / sec ) * 25 milliseconds or 3125000 (for read) * 40 milliseconds or 5000000 (for write) net.core.rmem_max = 16777216 net.core.wmem_max = 16777216 net.ipv4.tcp_wmem = 4096 3125000 16777216 net.ipv4.tcp_rmem = 4096 5000000 16777216 The 25 and 40 milliseconds are my guess at roughly the faster and more middle ping times within the domestic US Internet. This just gets each connection started net.ipv4.tcp_moderate_rcvbuf = 1 net.ipv4.tcp_timestamps = 1 cause the IP stack to very quickly establish precise values for the transmit and receive TCP windows.

1 0

Guard flag flapping
by starlight.2015q3＠binnacle.cx 08 Aug '15

08 Aug '15

First, I am assuming you are running bare-metal on a system and not in a virtualized server--everything below is premised on that. Do not expect a virtual server or Linux container to perform well as a high- capacity Tor relay. It's possible to configure a high-performance VM, but this is an esoteric art and one is better off renting a small dedicated physical server than going that route. Your story of a relay setup that should measure fast by all apparent metrics but is given terrible rankings by BWauths is common this year. BWauths scripts are known to be buggy, though supposedly have been improved very recently. 'longclaw' just came back online with the "latest" code, but after starting out with a failure to measure 2000 relays two days ago, it's still running 1000 shy of the full population: https://consensus-health.torproject.org/#bwauthstatus Scroll down a little and you will see 'longclaw' is unique in voting 976 relays not-guard and 1709 relays not-fast. That seems a more serious issue than cold start glitching IMO, and is not impressive if that is what it really is. A fifth BWauth is said to be arriving soon and it is said that it will help. Your relays currently are measured thusly: greendream848 longclaw-w Bandwidth=1694 Measured=986 gabelmoo-w Bandwidth=1694 Measured=347 maatuska-w Bandwidth=1694 Measured=874 moria1 -w Bandwidth=1694 Measured=1550 spacequeen974 longclaw-w Bandwidth=1698 Measured=493 gabelmoo-w Bandwidth=1698 Measured=970 maatuska-w Bandwidth=1698 Measured=1930 moria1 -w Bandwidth=1698 Measured=2130 You can see future and past reports of these in https://collector.torproject.org/recent/relay-descriptors/votes/ https://collector.torproject.org/archive/relay-descriptors/votes/ where longclaw is 23D15D9. . . gabelmoo is ED03BB6. . . maatuska is 49015F7. . . moria1 is D586D18. . . That the measurements are all in the same ballpark does indicate that some subtle issue with the network and/or equipment may be at work and the BWauths may not be at fault. But many have complained that nothing they do seems to work. If the firewall is performing stateful packet inspection or any kind of DPI (deep packet inspection) disable that for all incoming and outgoing Tor traffic. It's all encrypted anyway so there's no point, and DPI can drag down performance big-time. The directory traffic is unencrypted but I've never heard of a firewall with stateful rules for the Tor directory protocol. If you can put the system directly on the public IP address with no firewall or local-rack router I recommend doing this. Just make sure iptables are set to protect login and other non-tor access. Either that or disable iptables and strip the server down so that nothing but the 'tor' process and 'ssh' are running, and configure 'ssh' to accept only certificate authentication (be sure to set and test the cert auth before applying the setting). Check for minimized listeners with lsof -Pn | fgrep LISTEN The email daemon should stay up to handle alarms, just be sure it listens on 127.0.0.1. Likewise anything else that is absolutely necessary. Use *Port and *Policy settings in torrc to lock down control and socks access to the daemon. One notable sysctl that matters for high-capacity relays is net.netfilter.nf_conntrack_checksum = 0 though having this enabled would not cause the current poor measurements. You should change this setting: net.ipv4.tcp_no_metrics_save = 1 turning this off was to work around a very- long-ago kernel bug that is fixed everywhere. Turning it on improves performance. You might try net.ipv4.tcp_wmem = 4096 250000 4194304 net.ipv4.tcp_rmem = 4096 375000 4194304 which will cause the congestion window to get to full size a bit quicker, and these net.core.somaxconn = 1024 net.core.netdev_max_backlog = 524288 net.ipv4.tcp_slow_start_after_idle = 0 net.ipv4.tcp_keepalive_time = 600 which increase various limits for fast networks, lots of connections. Make sure these defaults values are active and have not been changed to non-default by /etc/sysctl.conf: net.ipv4.tcp_moderate_rcvbuf = 1 net.ipv4.tcp_timestamps = 1 net.ipv4.tcp_window_scaling = 1 net.ipv4.tcp_sack = 1 net.ipv4.tcp_syncookies = 1 net.ipv4.tcp_congestion_control = cubic And try adding TXQUEUELEN=100000 to the /etc/sysconfig/network-scripts/ifcfg-ethX for the interface(s) where tor runs. Manually activated with ip link set qlen 100000 dev ethX ip link show dev ethX Finally make sure the kernel is of a vintage with the Google-advocated connection-start congestion-window increase: https://lwn.net/Articles/427104/ http://samsaffron.com/archive/2012/03/01/why-upgrading-your-linux-kernel-wi… http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdi… If you end up implementing any of the above and it works please describe the results in tor-relays post.

1 0