Silk Road forums
Discussion => Security => Topic started by: SR Discount Pharmacy on July 30, 2011, 11:10 pm
-
I know TOR alone is bullshit. However, is TOR+ Double VPN enough? or do you guys suggest TOR + Double VPN + RDP. Right now I am on Tor + Double VPN and RDP and it is slowwww as you'd expect with routing your packets in 4 different places before connecting to Silk Road. I am wondering if perhaps TOR + Double VPN is enough? Or maybe just TOR+RDP?
-
fuck you must grind your teeth constantly while waiting for pages to load. I would be smoking two packs a day with that...
-
If you are located near a campus with a public wifi - that would be ideal. You could get an antenna to increase the range.
-
I know TOR alone is bullshit.
You've been misinformed.
Right now I am on Tor + Double VPN and RDP
Please don't do this.
You are mixing up threat models, and the result tends to be the inverse square of the number of security features you are adding.
Adding VPN to TOR = 1/4 the security of TOR alone. Adding a second VPN = 1/9th...
Either tonight or tomorrow I'll try to find the time to do a brief write-up about why VPN's dramatically lower your security on TOR.
-
Hmm can anybody else confirm this? VPNs reduce the anonymity of Tor? Even if they are logless? I have been practicing online anonymity for about 10 years now. The general school of thought is to chain things and that makes it more secure due to if one place is logging then nobody will have your real IP. (i.e VPN+Proxy).
Even if this is true, i am on anonymous 3g wireless bought with prepaid plastic which negates any problem with my real IP leaking.
The reason i said TOR is bullshit is because of the "evil exit nodes" problem. I didnt mean to say its bullshit. Its a great system, however i dont beleive it is enough for a seller. What if the DEA makes an exit node that looks at all the data? They get what the seller is saying and even his real IP. Then what comes next is easy to figure out. (warrants,raids,arrests)
I would really appreciate if you could send me that info....or if some other experts could chime in.
NOTE TO ALL SELLERS: Always use anonymous prepaid 3G Wireless bought with prepaid plastic. Preferably away from where you live. Do not do this stuff from home people. If people are interested in this ill type up a guide for obtaining an anonymous prepaid 3g card for further anonymity.
-
If people are interested in this ill type up a guide for obtaining an anonymous prepaid 3g card for further anonymity.
please do.
-
SR Discount Pharmacy, let me handle some of this a bit out of the order you brought it up in, so I can save the long-assed shit for last.
i am on anonymous 3g wireless bought with prepaid plastic which negates any problem with my real IP leaking.
Ever since CALEA, obtaining pretty darned accurate location data (i.e. within 20 yard) from a cellular device is trivial, no warrant required. If you're sitting at home expecting that to offer an extra layer of protection, forget it. If any agency is to the point where they know the cellular device you're using, it's only a matter of time.
The reason i said TOR is bullshit is because of the "evil exit nodes" problem. I didnt mean to say its bullshit. Its a great system, however i dont beleive it is enough for a seller. What if the DEA makes an exit node that looks at all the data? They get what the seller is saying and even his real IP. Then what comes next is easy to figure out. (warrants,raids,arrests)
Any exit node should be considered malicious.
TOR encrypts traffic to the surface web from your computer, through several hops, and right up to the point it hits the exit node. If it is non-SSL traffic, it is sent unencrypted from the exit node to the website, and ANY exit node can sniff the contents.
However, while malicious exit nodes deserve a long explanation in their own right, the short version is that when you are using TOR to access an .onion site, you don't have to worry about them. Traffic to an .onion site is encrypted end-to-end, and there is no 'exit' node as your traffic never leaves the TOR protocol.
Understand how TOR works.
First, to build a TOR circuit when you first start TOR, you are connected to one of two of your Guard Nodes. For a simple explanation, these Guard Nodes never change. Whenever you fire up TOR, it always seeks out the same Guard Nodes. If you check your TOR log during the initialization process, you will see where it 'guesses' your current IP address, according to another IP address. This is it 'pinging' one of your guard nodes, and it responds with the IP address it believes you are on. This sets up the first part of your secure circuit.
The next step, is a Relay Node.
Generally speaking again, you will get a different Relay Node every session, and the Relay Node has no idea where the fuck you are, only that the Guard Node is passing some encrypted traffic to it for you. If traffic varies wildly, or your Relay Node goes down or is swamped, it may change, but for the most part it's yours for as long as you are using the current session. The Relay Node cannot read your traffic, nor does it know where the final destination is. All that it does is pass the encrypted traffic along to an Exit Node.
The Exit Node is the only node that knows where your traffic is going, and if it's going to an .onion site, it DOES NOT, in fact CAN NOT decrypt it, all it can do is pass it on to the final site. Furthermore, it doesn't even know WHAT .onion site it is passing the information along to. In this case, it believes it is acting as a Relay Node, and not an Exit Node. So there can be no malicious exit node behaviour when accessing an .onion site.
By design, your exit node changes approximately every 10 minutes, whether you are visiting a .onion site or a surface web site, or even combinations of both at the same time. Even when surfing in combination, you only have one exit node.
Now, let's get on to the big question.
The general school of thought is to chain things and that makes it more secure due to if one place is logging then nobody will have your real IP. (i.e VPN+Proxy).
This is a lot easier to explain in a conversation than it is to put down in text and hope to be understood, but I'll give it a try.
Threat models.
Firstly, understand the threat model you are protecting yourself from.
If, say, I'm torrenting the shit outta every hit movie and song from the last twenty years, then my threat model tells me that the *IAA goons are going to try and track me down and extort money from me.
So, I fork over a few bucks a month to BTGuard or Ivacy and get a ridiculously high-speed vpn connection that doesn't log access, and I can torrent away to my hearts content. The *IAA goons have the VPNs IP address, but they say, hey, 'safe harbour DMCA provisions' that say they can't be held responsible for traffic over their network if they don't have knowledge of the content. They also don't have to keep logs, so the *IAA guys can go pound salt up their asses.
Perfect, right?
Yeah, if you're worried about getting sued for torrenting.
New threat model.
I'm doing some illegal shit, and there are some three letter agencies that would be interested in it. When you're talking drugs, they are all interested. Pick one - DEA, FBI, CIA, NSA...
Keep in mind the DEA has a budget of over a BILLION US$ A WEEK, and the NSA probably spends more on pencils in a week than the whole DEA budget. And the FBI never gave up on Carnivore.
So, you have one or more of your TLA's merrily monitoring every scrap of traffic on the internet, and that's not science fiction. 10 days of the DEA budget a year would buy it the storage capacity for every byte that moves accross the country, and you'd better believe the NSA doesn't just hijack ATT backbones, they copy traffic on every undersea fibre optic cable out there.
The good news is most that traffic is crap. Shitloads is encrypted, lots is broken up between several paths during the trip, and makes no sense in pieces, and the very vast majority of it is, to use the technical term, boring as shit to any of these TLA's.
So they look for anomalies.
Chaining VPNs is an anomaly.
The software asks itself, why is this traffic leaving a VPN, only to hit another VPN. Or, why is this traffic leaving a VPN, only to start it's way on a TOR circuit. (TOR relays are known, btw - a for the most part complete list of them can be found here: hxxps://metrics.torproject.org/networkstatus.html)
See, if you're doing that, they're pretty sure you're up to what's technically known as 'no good'.
And these guys at those TLAs aren't the goons from the *IAA.
They have WAY more resources, and aren't limited to doing things like asking for logs.
So they do what's called traffic analysis.
They compare the size, speed and frequency of the traffic from VPN to VPN, or VPN to TOR, that's coming out of the VPN, with the traffic that's going IN to the VPN. Now, by nature, your traffic to the VPN is 'tunnelled', that is to say encrypted. But it is designed to keep the information travelling back and forth secure, not the source and destination of that traffic.
Traffic analysis like this is what's known as a side-channel attack, and are very effective.
Everytime traffic leaves the VPN for TOR, they compare the traffic going in to the VPN for size, speed, etc. After a very short while, they can say with a great deal of accuracy, that the traffic leaving that VPN for another VPN, or for TOR, initiates at your IP address.
Now, while they still don't know WHAT you're doing, they're pretty sure it's 'no good' by their definition.
You've now got their attention.
And they know your IP.
They can even compare your outgoing traffic and use traffic analysis to correlate that with posts you make on an .onion site, if they decide to put the resources into it.
So, to put it in the jargon, using a VPN is contra-indicated for the threat model you are facing.
Now, set up TOR, and always runs as a relay, btw, and DO NOT add those extra layers of complexity that actually expose more than they hide, and you're golden. Unlike VPNs, which are designed to provide SECURITY for the data in transit, TOR is designed to offer ANONYMITY to you, and as a side effect also secures your traffic end to end when visiting an .onion site.
-
what he said.
-
So they look for anomalies.
Chaining VPNs is an anomaly.
The software asks itself, why is this traffic leaving a VPN, only to hit another VPN. Or, why is this traffic leaving a VPN, only to start it's way on a TOR circuit. (TOR relays are known, btw - a for the most part complete list of them can be found here: hxxps://metrics.torproject.org/networkstatus.html)
See, if you're doing that, they're pretty sure you're up to what's technically known as 'no good'.
And these guys at those TLAs aren't the goons from the *IAA.
They have WAY more resources, and aren't limited to doing things like asking for logs.
So they do what's called traffic analysis.
They compare the size, speed and frequency of the traffic from VPN to VPN, or VPN to TOR, that's coming out of the VPN, with the traffic that's going IN to the VPN. Now, by nature, your traffic to the VPN is 'tunnelled', that is to say encrypted. But it is designed to keep the information travelling back and forth secure, not the source and destination of that traffic.
Traffic analysis like this is what's known as a side-channel attack, and are very effective.
Everytime traffic leaves the VPN for TOR, they compare the traffic going in to the VPN for size, speed, etc. After a very short while, they can say with a great deal of accuracy, that the traffic leaving that VPN for another VPN, or for TOR, initiates at your IP address.
Now, while they still don't know WHAT you're doing, they're pretty sure it's 'no good' by their definition.
You've now got their attention.
And they know your IP.
They can even compare your outgoing traffic and use traffic analysis to correlate that with posts you make on an .onion site, if they decide to put the resources into it.
So, to put it in the jargon, using a VPN is contra-indicated for the threat model you are facing.
Now, set up TOR, and always runs as a relay, btw, and DO NOT add those extra layers of complexity that actually expose more than they hide, and you're golden. Unlike VPNs, which are designed to provide SECURITY for the data in transit, TOR is designed to offer ANONYMITY to you, and as a side effect also secures your traffic end to end when visiting an .onion site.
How would you even have this information, to this extent?
-
Paraphrase:
"they know your IP"
How is it that they know your IP? They should be able at most to obtain the shared IP of the VPN provider. They would then have to access the VPN provider to get the records (that are not kept) to ascertain your real IP. Restarting your VPN session every so often (20minutes?) would change that IP too.
"traffic analysis correlations"
Very True but who says that the only traffic you run TO the VPN is tor? Personally My other browser or programs (video streaming, bitttorent) are running outside of my sandboxed and tor'd Tor browser.
Traffic correlation to what goes "IN" to the VPN would not correlate in the least. If I have 200kb of channel data transmitting to my VPN and tor makes up maybe 15kb per sec of that data intermittently how does an attacker segment total encrypted traffic volume to "end" tor site or onion site beyond a large degree of uncertainty? How does this not get lost in the other VPN/Tor traffic crowd browsing similar sites?
Relaying traffic in addition is a no brainer.
(side note: If traffic analysis is so powerful then TOR should add a random "time salt" to relays which switches routing order (via cache) or ads a slight time delay. I know I would not care about .8ms to 1 second chained delays to increase anonymity. Easier if that setting could be set by a flag in the client itself...increasing anonymity by defending against time analysis for those who need it the most)
"Double VPN vs anonymity"
I think chaining through different services certainly could make your traffic stand out and reduce anonymity. But if chaining within a service where it never exits the the real web in the chain (ivacy) an attacker would have to be internal to the VPN network to "see" that the traffic is chained. However and "evil honeypot" VPN could see the chained traffic but all of your traffic including tor would appear this way. To that VPN host you would look privacy concerned like most who use it but not suspicious.
Ultimately if you got some real "dirt" to do "borrowing" (with a cloned throw away mac address) a connection or using a public one (where nobody can peep your screen) before implementing any of your anonymity schema would be best. I do believe that "logging into" a vpn account in this type of situation could possibly reduce anonymity. If logs are truly not kept then maybe not but why use a trust model for dirt? So in public Tor with relay running...at HOME VPN so your ISP cannot see or record tor access during dirt giving you deniability.
At the end of the day if the "find" you they also need to be able to decrypt your hard drive and or find evidence of what you are doing/selling. Who keeps then ships their shit from home? Layer your fucking life! Make it difficult not easy to be caught. I even fill my orders from a different city...sometimes state. In time let an associate fill your large orders "in country" to avoid customs and you fulfill his/hers or just transfer the wholesale payment.
======
I am not a programmer or privacy expert but I've been doing dirt and staying private for @15 years now and believe a layered approach when running from "home" is best.
Never do the same thing in the same place all of the time. Change up.
-
TOR encrypts traffic to the surface web from your computer, through several hops, and right up to the point it hits the exit node. If it is non-SSL traffic, it is sent unencrypted from the exit node to the website, and ANY exit node can sniff the contents.
However, while malicious exit nodes deserve a long explanation in their own right, the short version is that when you are using TOR to access an .onion site, you don't have to worry about them. Traffic to an .onion site is encrypted end-to-end, and there is no 'exit' node as your traffic never leaves the TOR protocol.
True
First, to build a TOR circuit when you first start TOR, you are connected to one of two of your Guard Nodes. For a simple explanation, these Guard Nodes never change. Whenever you fire up TOR, it always seeks out the same Guard Nodes. If you check your TOR log during the initialization process, you will see where it 'guesses' your current IP address, according to another IP address. This is it 'pinging' one of your guard nodes, and it responds with the IP address it believes you are on. This sets up the first part of your secure circuit.
Wrong, Tor uses three guard nodes and they change roughly once every month. You create circuits with multiple guards simultaneously.
Generally speaking again, you will get a different Relay Node every session, and the Relay Node has no idea where the fuck you are, only that the Guard Node is passing some encrypted traffic to it for you. If traffic varies wildly, or your Relay Node goes down or is swamped, it may change, but for the most part it's yours for as long as you are using the current session. The Relay Node cannot read your traffic, nor does it know where the final destination is. All that it does is pass the encrypted traffic along to an Exit Node.
Mostly true, relays are selected from the entire list rather than a guard list so they do change a lot. Session lasts for ten minutes before new connections are sent through a new circuit.
The Exit Node is the only node that knows where your traffic is going, and if it's going to an .onion site, it DOES NOT, in fact CAN NOT decrypt it, all it can do is pass it on to the final site. Furthermore, it doesn't even know WHAT .onion site it is passing the information along to. In this case, it believes it is acting as a Relay Node, and not an Exit Node. So there can be no malicious exit node behaviour when accessing an .onion site.
True, you don't even need to use an exit node to exit to a hidden service.
By design, your exit node changes approximately every 10 minutes, whether you are visiting a .onion site or a surface web site, or even combinations of both at the same time. Even when surfing in combination, you only have one exit node.
Hidden services always use a new unique circuit.
I'm doing some illegal shit, and there are some three letter agencies that would be interested in it. When you're talking drugs, they are all interested. Pick one - DEA, FBI, CIA, NSA...
NSA doesn't give the slightest fuck about drugs. Furthermore, NSA can trivially trace servers through Tor and they wont have a hard time tracing all users of any website simply by using application layer traces instead of network layer. They have fuck tons of zero day exploits. They also can directly gather signals intelligence on Tor and deanonymize a fuck ton of its users in this way, other than a few lucky ones who pick just the right circuits. In short, NSA can trace you.
CIA I don't know as much about, they are still not at all likely to become involved with taking SR down though. They might not be as technically leet as FBI since their focus is human intelligence, managing networks of spies and informants, versus signals/communications intelligence, mapping out networks and gathering communication information from them. They are more likely to sell drugs via SR to fund some project they want to hide from the rest of the government. DEA and FBI are our likely adversaries for our threat model, not NSA and CIA.
Keep in mind the DEA has a budget of over a BILLION US$ A WEEK, and the NSA probably spends more on pencils in a week than the whole DEA budget. And the FBI never gave up on Carnivore.
Carnivore is easy to bypass. Also keep in mind that government agencies spend about $100,000 per pencil to their friends pencil company. That is part of the full government scam.
So, you have one or more of your TLA's merrily monitoring every scrap of traffic on the internet, and that's not science fiction. 10 days of the DEA budget a year would buy it the storage capacity for every byte that moves accross the country, and you'd better believe the NSA doesn't just hijack ATT backbones, they copy traffic on every undersea fibre optic cable out there.
I talk with a lot of people who are signals intelligence experts. Most of them think that even the NSA can only sample traffic at IXs, meaning they only log one out of every X packets. X is probably in the thousands. And this is only traffic passing through IXs in USA. Even the less conservative experts I talk with do not think that the NSA can log all traffic that is inside the USA, but they think they can log all traffic that enters or leaves the country. Thinking that the NSA can log 100% of all traffic on the internet is a minority opinion, although I have read one paper indicating that it is possible I am of the opinion that the paper is not correct based off of my extensive interaction with a wide range of experts and my own personal knowledge of signals intelligence.
Chaining VPNs is an anomaly.
It's at least not likely to be very helpful after a few hops, because its end to end correlation attacks to worry about the most not log trace attacks.
The software asks itself, why is this traffic leaving a VPN, only to hit another VPN. Or, why is this traffic leaving a VPN, only to start it's way on a TOR circuit. (TOR relays are known, btw - a for the most part complete list of them can be found here: hxxps://metrics.torproject.org/networkstatus.html)
See, if you're doing that, they're pretty sure you're up to what's technically known as 'no good'.
If they can see where the traffic exits one VPN and hits another VPN, and trace that back to you, then the first VPN solution has failed in the first place and it would have been better to use a second VPN on the chance that they can tell you are using two VPNs (which requires they trace the first VPN part for the YOU to hold true) PLUS trace the second VPN to link YOU to the site you are visiting. So chaining VPN actually does add more anonymity from log trace attacks. The attacker who does log trace attack is far more likely to be at the other end of your path when they first start trying to trace you, than at some encrypted point in the middle of your chain.
They have WAY more resources, and aren't limited to doing things like asking for logs.
So they do what's called traffic analysis.
Traffic analysis .... is the analysis of traffic .... which is recorded in the form of logs.
They compare the size, speed and frequency of the traffic from VPN to VPN, or VPN to TOR, that's coming out of the VPN, with the traffic that's going IN to the VPN. Now, by nature, your traffic to the VPN is 'tunnelled', that is to say encrypted. But it is designed to keep the information travelling back and forth secure, not the source and destination of that traffic.
This is called an end to end correlation attack and is traffic confirmation. To do an end to end correlation attack like this they need to be able to see the server you are communicating with. That will require in some cases obtaining logs from the server, in other cases the server tells them all they need to know. For example, this post is public and has timestamps on it so they don't need logs to see this information. Some researchers would call it a fingerprinting attack at this point rather than an end to end correlation attack. All low latency solutions are weak to correlation attacks, but as long as your entry guards are not bad you wont be susceptible to them.
Traffic analysis like this is what's known as a side-channel attack, and are very effective.
No this is not a side channel attack. Side channel attacks involve neutralizing the benefits of a solution with out directly attacking the solution. For example, java/flash side channeling Tor by going around it. Correlation attacks are a direct attack on Tor that it is weak to by design.
Everytime traffic leaves the VPN for TOR, they compare the traffic going in to the VPN for size, speed, etc. After a very short while, they can say with a great deal of accuracy, that the traffic leaving that VPN for another VPN, or for TOR, initiates at your IP address.
They can only do this if they already have you under surveillance. Traffic analysis = doing general surveillance of a network trying to link communicating parties together, or trying to find a specific target out of a crowd. Tor protects from traffic analysis. Traffic confirmation = confirming that a target communicates with a known server. Tor does not protect from traffic confirmation. Tor makes it so you are no more likely than your neighbor who uses Tor to be going to SR, but it does not prevent an attacker from analyzing your traffic and your neighbors traffic and traffic to SR and determining if you or your neighbor go here.
And they know your IP.
They need to already know your IP to do a traffic confirmation attack (show that you go to SR), and Tor protects them from knowing to target your IP because it protects from traffic analysis (analyzing Tor to see who all goes to SR).
They can even compare your outgoing traffic and use traffic analysis to correlate that with posts you make on an .onion site, if they decide to put the resources into it.
If any of your entry guards are owned, you are fucked. Tor and all low latency solutions are over rated. It all comes down to the entry guards. Thankfully FBI and DEA are not CIA and NSA and they don't know what the fuck they are doing. If DEA and FBI were actually skilled, they would be working at CIA and NSA. They are the retarded half brothers of the intelligence community.
Now, set up TOR, and always runs as a relay, btw, and DO NOT add those extra layers of complexity that actually expose more than they hide, and you're golden. Unlike VPNs, which are designed to provide SECURITY for the data in transit, TOR is designed to offer ANONYMITY to you, and as a side effect also secures your traffic end to end when visiting an .onion site.
Running as a relay hurts your anonymity in many cases and you should NOT run as a relay for our threat model. Running as a relay means that everyone knows you are a Tor user. And roughly where you are located. And they already know someone who is a Tor user roughly where you are located ships drugs. Use a bridge maybe.
-
@nsapwnsall-what would be the most complete and effective list of entry guards that you're referring to? This info would be most helpful to us who are waaaaay less knowledgeable than you and some of the other posters. Thanks! :)
-
Hey nsapwnsall, thanks for chipping in. I hope you are going to stick around and continue to participate in the discussion.
ButtPlug got us thinking over here today, and we've spent the better part of the day running TOR traffic through ivacy and getting some metrics. We ran a series of tests concerning timing correlation attacks, and some of the results (like when running a torrent at the same time) actually were kind of surprising.
We're just wrapping up for the day now, and tomorrow I'll post up some of our base raw data and analysis of it. I'm hoping nsapwnsall checks in on it as well.
Edit @ nsapwnsall: I'll respond later to a couple of your points, as I think you misinterpreted a bit of my rushed post earlier. However, I am *pissed* to discover not only that you are correct that it's 3 guard nodes, not 2, but that the change was made 4 years ago. The last time I delved deeply into the workings of TOR was at Blackhat '07, at which time TOR used 2 guard nodes. Thanks for correcting me.
Just goes to show that I've gotta keep up.
-
Damn i learned so much from this thread. There are some seriously smart people on this board. So to sum it up i would conclude the best anonymity would come from:
Tor + Anon 3g wireless for hidden services/.onion
VPN+Socks5 or Double VPN+Socks5 for Clearnet
-------------------------------------------------------
Shabang i cant wait to see your results bro
-
As a much more novice person to all of these things I would be VERY impressed if someone gave a complete listing of what needs to be done:
1. to not be located geographically while running Tor. Presuming most people are going to be at home using a cable modem. Or if that's just preposterous then what is the next most feasible, both in terms of convenience and installation. IF the answer ends up being that I have to be at some university library using their system, well, I'm going to shoot myself. ;)
2. the best setup on your own machine-which appears to be running TAILS off a flash drive with a second flash drive for your files, etc.
If I'm full of crap or off on a way wrong tangent, I'd love for someone to both set me straight and fill in the blanks here. :) much obliged!
-
Damn i learned so much from this thread. There are some seriously smart people on this board. So to sum it up i would conclude the best anonymity would come from:
Tor + Anon 3g wireless (prepaid 3g not from home) for hidden services/.onion
VPN+Socks5 or Double VPN+Socks5 (prepaid 3g not from home) for Clearnet
That is the only way to be 100% safe in my opinion.
-------------------------------------------------------
Shabang i cant wait to see your results bro
-
I made a post about tails in the security section... waiting for the page to load and I'll link it.
It is entirely possible to have tails + encrypted storage on the same flash device.
Here is a semi-tutorial. Should be good enough to figure it out, but feel free to ask anything if you try it.
http://dkn255hz262ypmii.onion/index.php?topic=1382.0
I used a uSD card, and it worked fine. (should be more secure)
"2. the best setup on your own machine-which appears to be running TAILS off a flash drive with a second flash drive for your files, etc."
-
I read that using TAILS can make you weak against profiling attacks, because it doesn't save guard nodes.
-
First, to build a TOR circuit when you first start TOR, you are connected to one of two of your Guard Nodes. For a simple explanation, these Guard Nodes never change. Whenever you fire up TOR, it always seeks out the same Guard Nodes. If you check your TOR log during the initialization process, you will see where it 'guesses' your current IP address, according to another IP address. This is it 'pinging' one of your guard nodes, and it responds with the IP address it believes you are on. This sets up the first part of your secure circuit.
Wrong. The guard you're using for circuit A may not be the same as the guard you are using for circuit B. Tor will choose the guard from a set of three relays. Over time, the relays in that set may change for various reasons.
The Exit Node is the only node that knows where your traffic is going, and if it's going to an .onion site, it DOES NOT, in fact CAN NOT decrypt it, all it can do is pass it on to the final site. Furthermore, it doesn't even know WHAT .onion site it is passing the information along to. In this case, it believes it is acting as a Relay Node, and not an Exit Node. So there can be no malicious exit node behaviour when accessing an .onion site.
Just want to point out that you have more than three hops when visiting a hidden service, as explained on https://www.torproject.org/docs/hidden-services.html.en
By design, your exit node changes approximately every 10 minutes, whether you are visiting a .onion site or a surface web site, or even combinations of both at the same time. Even when surfing in combination, you only have one exit node.
The whole circuit changes every 10 minutes, not just the exit relay.
-
If you are located near a campus with a public wifi - that would be ideal. You could get an antenna to increase the range.
..yes ideal, however access might not last forever nor may be relaible/within range signall etc.. plus security is supposed to/usually implemented to prevent abuse...but by all means give it a try.
in another thread i recommended piggy backing someones home wifi w/o them knowing, i wouldnt normally do this/recommend this but if its for SR you are forgiven ...hahaha :P
{again may be locked down and / or require the access key, but lot of users are not techno savy and may be wide open..}