For a program I am writing, I would like to use TLS (or something similar) to encapsulate my application's protocol. This will minimize both the amount of work I have to do as well as the number of vulnerabilities I could accidentally create.
My program is designed to be peer-to-peer although one or more servers provide some services to help one user locate another (it registers IP address/port combos) but do little else. I want to make this system very fault-tolerant so having these servers act as a Certificate Authority is unacceptable because a compromise of a server or its key would affect too many users. Therefore I plan on using a Web of Trust.
The main problem with using TLS is that the original TLS 1.2 specification (RFC 5246) does not provide for using OpenPGP certificates. It seems to be very x.509 centric. RFC 6091, which obsoletes RFC 5081 and extends RFC 5246, makes provisions for an extension to TLS that does what I want. The problem is that I don't think BouncyCastle implements this extension and I can't find a Java crypto library that does. I also don't want to write my own / contribute to BC because I'm really bad at not making mistakes and I'm also very lazy.
Another problem with this is that BouncyCastle provides "a light weight client-side TLS API" but because this software is P2P, a server-side API is also necessary so that I can use TLS by making it believe that the peer originating the connection is the client. I'm pretty sure that once the handshake is complete that it's the same.
Questions:
Is there any way that I can still use TLS (which I highly doubt)? Is there a protocol like TLS that is designed for P2P, or at least can function in this way (like I believe TLS can), but can work with an OpenPGP certificate? If neither is the case, should I pursue the idea explained in this question and implement my own layer taking concepts from TLS?
Links to RFCs: RFC 5246 and RFC 6091
The only library that I know to support RFC 6091 (i.e. TLS with openpgp certificates) is GnuTLS but I don't know whether you can use something like that in Java. Alternatively you could replicate the SSH semantics, where you store the public keys of your peers using self-signed
X.509 certificates.
In TLS, the X.509 parts are actually handled as opaque blobs:
The server sends its certificate (and some helper certificates, if it wishes so) as (a list of) opaque string(s) of bytes (a three-byte length, followed by the encoded certificate as arbitrary bytes).
When the server asks for public key client authentication, it sends a list of "names" which are supposed to be the encoded X.500 names of the root CA the server will recognize -- there again, opaque blobs (two-byte length).
The client, when (if) it sends a certificate (chain), uses the same format than the server.
As TLS is defined, both client and server are supposed to use the peer public key, which they get in any way they see fit and that's mostly out of scope of the TLS specification: the certificates exchanged over the wire are considered as mere helpers. So there would be no problem in actually sending OpenPGP encoded public keys in those blobs, as long as both client and server expect it -- and since you control code on both, this should be no issue.
Your problem then "simply" becomes a matter of making a TLS implementation accept to hand you the blobs without choking on them. I know of no existing Java-only TLS implementation which will fit the bill, so you may have to write a bit of code -- but I urge you not to fiddle with TLS protocol details except processing of the certificate blobs. Those things are subtle and weaknesses are sooo easy to create...
As far as I know, the Sun/Oracle JSSE implementation only deals with X.509 TrustManagers (which you can customize to handle certain extensions, but would still expect a structurally valid X.509 certificate.
It might be possible to use Java's security API to implement RFC 6091, but I'm not sure how. It's definitely more work than just tweaking the TrustManagers, as you would have to go deeper into Java's implementation of TLS.
Alternatively, if it's for a bespoke server, you could re-use the key material from PGP certificates into X.509 certificates and put the initial PGP certificate (with all its signatures) as a blob in a custom X.509 extension (as it's more or less done here). The problem here would be interoperability, since such an extension wouldn't be a standard. Implementing a TrustManager in Java that is able to understand extension is definitely feasible, and you wouldn't need to dig into the internals of Java's TLS stack, you'd only have to deal with custom TrustManagers to initialize your SSLContexts.
Related
Currently the TLS Session Resumption documentation for Node.js only covers two methods (Session IDs and Session Tickets) both of which are obsoleted in the TLS 1.3 spec. In practice I've found that Filezilla is not utilizing either of these methods over TLS 1.3 when the server is configured to handle them.
That was confirmed to me by a Filezilla team member, they only utilize the Pre-Shared Key method as specified in TLS 1.3.
Which brought me to the PSK docs, which are sadly even more sparse and dated than the session resumption docs. They don't even cover exchanging a PSK during the handshake and instead focus on out of band PSK. I've been digging through the Node source but am not finding anything very helpful since most of this stuff is in the deps -> OpenSSL and my skills at deciphering C are not great.
It is quite possible that my question is not suitable for this medium, if so I understand if it is closed. That said, my question is simply: Can PSK based session resumption be achieved in Node.js at all at this point and if so, what would that flow look like?
Background: I'm working on a mobile application (online chat) that use persistent TCP connections to a backend server. In the previous version, we used a well-known encryption scheme to protect the payload and sign using a shared secret key. This provided sufficient protection against data tampering and reverse engineering. But the traffic was still susceptible to replay attacks: someone can capture the network traffic and replay it and server would accept the requests since it had no idea whether it was real or replayed. To counter this, we introduced timestamps and nonces which provided protection against replay attacks.
All well and good. A decision was later made to use TLS (Transport Layer Security) to protect all communication between clients and servers. TLS was implemented and to provide an extra layer of protection against Man-in-the-Middle attacks, we pinned the public certificate of the server on clients.
Which brings me to the question: is it still necessary to use custom encryption and guard against replay attacks since TLS already provides protection against these? In my research, I found that TLS provides protection during transmission i.e. prevents man in the middle but what if the originator itself wants to cheat the system and modify payload or replay traffic? Does TLS protect against an intentional cheater trying to somehow intercept the traffic in the network layer of their OS? My understanding was that TLS provides end-to-end encryption but I have a suspicion that TLS traffic could, in fact, be replayed if the originator wishes to.
I'm also aware of memory-based attacks where the attacker would modify the memory on the device this modifying payload (may be even record and replay.) But even our custom encryption + replay guard won't protect against these.
From: https://www.owasp.org/index.php/Transport_Layer_Protection_Cheat_Sheet#Benefits
"TLS also provides two additional benefits that are commonly overlooked; integrity guarantees and replay prevention. A TLS stream of communication contains built-in controls to prevent tampering with any portion of the encrypted data. In addition, controls are also built-in to prevent a captured stream of TLS data from being replayed at a later time.
It should be noted that TLS provides the above guarantees to data during transmission. TLS does not offer any of these security benefits to data that is at rest. Therefore appropriate security controls must be added to protect data while at rest within the application or within data stores."
TLS only protects the transport and thus it provides protection against modifying or replaying of the encrypted data only. It does not protect against any kind of modifications or replaying of the data before the encryption or after decryption. Sending the same data again over a TLS connection is actually perfectly valid.
But, the nonce and timestamp you use to detect replay do not protect against modification or replaying too. The sender can still use the same data but "protect" these with the new nonce and the new timestamp. It is simply not possible to protect against this kind of client based attacks as long as the attacker is able to modify the client code which is usually the case if the client runs on a system owned by the attacker like in case of a smartphone game where the owner of the smartphone likes to cheat.
I'm working on a network program and I don't want anyone to know what kind of information is being passed when they sniff the network. Would using TLS achieve this? My main reason is that I want to keep the protocol I'm using to myself for now. If not please tell me if there is anything that can achieve my goal.
It depends on a lot of things, e.g. what your exact threat model is, and how much information leakage you can tolerate.
For TLS to provide adequate protection, these assumptions must be true:
Obviously, you should use a correct implementation, otherwise, if for instance, you are using SecureTransport from iOS 7.0.4, all bets are off.
You should enforce a minimum version requirement and only support secure ciphersuites. If you allow downgrade to SSLv2, you are setting yourself up for problems.
You check for validity of the server public key. You'd be surprised how many client apps skip this.
You use client certificates to authenticate the client, as well as the server, otherwise, it is possible to write a phony client that talks to your TLS server and reverse engineer your protocol. (You can also authenticate the client early in the protocol lifecycle using other means, but that part of your protocol would not be safe).
You keep the private keys secure.
(If you are using X509 certificates and trust chains:) Certificate authorities that you trust do what they are supposed to do, i.e. not sign certificates in your name for others.
You will still leak some packet length and timing information that you hope would not be complete enough for the reverse engineer.
The attacker does not control your client or server or have access to the binaries on any side. If, like an iPhone app, you are giving away the client binary, you have already lost.
Your higher level protocol cannot be tricked into say, redirecting to another server blindly, or lose its mind and do some other crazy thing when the client secure channel is interrupted. This can be hard to notice at times and depends on many other factors.
Something else I have probably missed here.
Would TLS prevent others reverse engineer my protocol?
Probably not. Pentesters do it all the time. They use something like Burp Suite to proxy the connection and watch all the web requests.
If not please tell me if there is anything that can achieve my goal.
Common practice is: if you don't want it stolen, copied, pilfered, abused, etc, then you don't put it on a client. So all sensitive code and data goes on a server you control. Since the client gets to see the request, you have to remove all sensitive information from it.
I'm in the designing stages of a custom tcp/ip protocol for mobile client-server communication. When not required (data is not sensitive), I'd like to avoid using SSL for overhead reasons (both in handshake latency and conserving cycles).
My question is, what is the best practices way of transmitting authentication information over an unencrypted connection?
Currently, I'm liking SRP or J-PAKE (they generate secure session tokens, are hash/salt friendly, and allow kicking into TLS when necessary), which I believe are both implemented in OpenSSL. However, I am a bit wary since I don't see many people using these algorithms for this purpose. Would also appreciate pointers to any materials discussing this topic in general, since I had trouble finding any.
Edit
Perhaps the question should have been: is there a best practices approach for secure passwords over unencrypted tcp/ip? If not, what are the reasons for selecting a particular method over others? (The Rooks answer is closest in spirit to this question so far, even if it does violate the letter).
Edit, part deux
I'm primarily interested in the case of client-server authentication, where there is an expectation that both parties have a shared secret (password) a priori.
You should have a look at "Diffie-Hellman key exchange":
Diffie–Hellman key exchange (D–H) is a cryptographic protocol that allows two parties that have no prior knowledge of each other to jointly establish a shared secret key over an insecure communications channel. This key can then be used to encrypt subsequent communications using a symmetric key cipher.
Once you have exchanged a key, you can encrypt your password with this key and transmit it over the insecure protocol.
I still think that SSL is by far your best choice, after all why reinvent the wheal when so much can go wrong? You don't have to buy an expensive certificate if your have a list of "good" and "bad" (compromised) certificates. openSSL is completely free, and i don't see a good reason not to use it.
Some things you might not know: ssl handshakes can be resumed.
Also you can use SSL/TLS over UDP to reduce overhead its called DTLS.
You could use a challenge-response algorithm. The algorithm goes like this:
The server sends a random string to the client.
The client combines this string with the password (by combining, you can xor them or just append them).
The client calculates a hash (for example, SHA1) of the result, and sends it to the server.
The server calculates the same hash using this random number and the real password.
The server compares the two hashes.
Since you shouldn't store a password in plain text, but as a hash instead, the client should calculate this hash at the very beginning.
There are possibly several libraries implementing this, so you probably don't need to code it yourself.
I'm using RSA to encrypt communication between a server and a client.
Lets say we have 2 Asymetric keys, key 1 and key2.
The server has key1 (Private) from the start and the client has the key1(public)
So here is the scenario:
the client generates key2
client connects to the server
sending key2(public) encrypted with key1(public)
from now on the server will send all data encrypted with the key2(public)
the client sends some random data to the server
the server sends back the same data hashed
the client verifies that the data is right
As far as I can see this should prevent a man-in-the-middle attack, or am I missing something?
At point 7 the client should know if someone is trying to give the server the wrong key to encrypt with, as no one else but the server can decrypt key2(public).
If there is anything that can be done to improve the security please tell me.
The best thing you can do to improve the security is to use an existing design and not try to reinvent the wheel. I'm not saying that what you've done is necessarily wrong, but just that many people much smarter than you and me have spent a lot of time thinking about this problem. Use TLS instead.
As long as key1 (private) has not been intercepted somehow by a third-party, your scenario looks secure.
I think I saw this somewhere in a paper actually. In it, Alice gave Bob an unlocked box (key 1 public), then Bob put a bunch of his own boxes (key 2 public) in it, locks it and sends it back to Alice. Alice then opens the box(key 1 private), and now she can securely seal the boxes that Bob just gave her.
Despite the box analogy, that's essentially what you're doing, so I'd say its secure.
I agree, just use TLS.
Also, what value do steps 5 through 7 provide? A MITM wanting to do an attack that would work after steps 1-4 (e.g. DoS of some sort by passing n transactions through and then stopping, forcing a retry from the start) could do so just as well after 5-7. What do they add?
-- MarkusQ
No, this protocol is not safe.
A man-in-the-middle can intercept the data sent by the client and send whatever it wants to the server, since you haven't specified any mechanism for the server to authenticate the client or verify the integrity of messages it receives.
Sure, you could doctor up your protocol to fix these glaring problems, but there would be others. If you ever fix them all, you'd have something that maps to TLS or SSH, so why not just start there?
#Petoj—the problem I was focusing on was that of the server trusting the messages it receives; your proposal doesn't provide any security there. However, if you are worried about confidentiality, you still have a problem, because the MITM could pass messages back and forth unaltered until he sees what wants to find because you don't have any privacy on the client messages.
Your proposal seems to be aimed at ensuring the integrity of messages from the client. You've developed the protocol to the point where the client can't distinguish between an attack and a network failure. Rather than trying to help the client determine whether the server acted on a tampered message, allow the server to verify the integrity of the message before acting on it.
I will agree with Greg that you are reinventing the wheel. What you are essentially describing is some basic form of key exchange. Incidentally, in order to ensure that it is secure against man-in-the-middle attacks you must also be certain of the server's identity, i.e. ensure that the client can know with certainty that what it believes to be public(key1) really is the server's and not the man-in-the-middle's (e.g. using a CA or having the server's public(key1) in secure storage on the client side.)
Moreover, there are additional considerations you must be aware from a systems standpoint, such as:
asymmetric key encryption is slower than symmetric key encryption, which is one of the reasons why existing solutions such as TLS will use asymmetric key encryption only to negotiate a temporary symmetric key, which is then used for channel encryption.
if traffic analysis by a third-party succeeds in cracking a temporary symmetric key, you have not compromised you asymmetric key pair. You are encouraged to re-negotiate the temporary key relatively often for this reason. Arguably, generating a new key2 in your scenario would mitigate this aspect.