I'm having trouble blocking the use of HTTP/2 in order to force browsers to use HTTP/1 as the protocol inside https. TLS MITM is out of the question, NFQUEUE-like usermode packet filtering may be considered at most, depending on processing overhead.
From reading the ALPN RFC it's not clear if providing an alert response or dropping the connection when I see a ClientHello that contains ALPN:http/2 will make the browser retry without ALPN.
If I understand correctly, modifying the ClientHello is out of the question as it results in a checksum error when the server responds with the ServerHello because changing the ClientHello invalidates the MAC for that packet.
Is the behavior for blocked handshakes w.r.t. ALPN the same as for handling TLS version fallback, i.e. TLS_FALLBACK_SCSV ?
EDIT: according to t1_lib.c from openssl, if the server doesn't know about ALPN, it ignores it. So if the server returns Alert for a ClientHello containing ALPN, it's probably only because it doesn't support TLS1.2, there is no way to signal to the client "please retry without ALPN" other than "alert" which results in the client trying TLS1.1.
HTTP/2 over TLS is negotiated via ALPN.
Browsers will tell to servers that they support it.
If you don't want to use HTTP/2, then you just have to modify the server configuration in a way that it doesn't have h2 as one of the protocols that it can negotiate via ALPN.
The ALPN negotiation will then fall back to HTTP/1.1 and the client will use HTTP/1.1.
Related
I've written a MIDlet server that handle tcp inputstream. I send xmlhttprequest with an http:// url from a html web page and it works fine. I want to do the same with an https url.
with my secureserverconnection instancied by a connector.open the accecptandopen() method return an exception.
handshake client send
As you can see the xmlhttprequest send an handshake hello but I've notices tls1.2 protocol then tls1.0 protocol and finally tls1.2.
Is it normal having 2 different protocol sent by the client. It is the first phase of the handshake protocol.
My Midlet java server respond with an Alert and a a "ssl3_get_client_hello:no shared cipher" error (IOException)
IOexception capture
I've specified the connector options like this : cipher : TLS_RSA_WITH_AES_128_GCM_SHA256 and TLSv1.2 protocol
tlsv1.2 server response
Have someone ever done it or can explain me why the cipher handshake is in error?
Yes it is normal for the client to try both protocols. The most likely scenario is that your device simply lacks any valid certificates, since J2ME devices are so old.
See:
"ssl3_get_client_hello:no shared cipher" in server depending on server certificate and key
I'm working on an academic project about TLS handshakes and i have captured some TLS traffic generated by multiple clients (google chrome, firefox...) and I want to see if for a given browser the client hello message will always be the same or no (I have removed the GREASE extensions because they are added in a random way to the client hello message and I omitted the SNI). I found that the same browser generate multiple client hello messages.
Is it normal to see such behavior or I'm doing something wrong?
A TLS handshake is done for each TCP connection involved in HTTPS and it is common that the browser uses multiple TCP connections in parallel. This is probably what you see. Multiple TLS handshakes within the same TCP connection are uncommon but might happen if a server requires a client certificate only for a specific path and thus triggers a renegotiation.
gRPC is based on http2 which must be use ssl.
But I found I can create use insecure server.
I want to know insecure means that don't use ssl or use predefined certificate for http2?
HTTP/2 does not insist on SSL/TLS. But all web browsers only implement HTTP2 over SSL/TLS because of problems when using it over plaintext HTTP over the Internet. So if not using a web browser but another HTTP/2 client than you can use HTTP/2 without SSL/TLS.
So yes insecure gRPC is using an unencrypted HTTP/2 connection (h2c). Though some implementations do not support this, similar to how web browsers do not support unencrypted HTTP/2 (h2c).
A client installed on jBOSS is trying to access a secured website configured on DataPower xi50v6.0.0.2 appliance. The connection is getting failed at SSL handshake.
I have taken a packet capture at DataPower and observed that SSL Handshake is failing with the Description:Handshake failure(40).
However, at the Client Hello step, I have observed that, only one Cipher Suite is specified which is : TLS_EMPTY_RENEGOTIATION_INFO_SCSV.
The TLS protocol used ( as per packet capture) is TLS1.1. Can this Cipher Suite be a problem?
In the DataPower system logs I can see below error:
Request processing failed: Connection terminated before request headers read because of the connection error occurs
Update:
The client application is running on jBOSS7.I have asked our jBOSS administrator to check the configuration at jBOSS end. I somehow got the access to server where jBOSS instance is installed and checked domain.xml where the ssl is configured. Where exactly in domain.xml, ths configuration related to cipher suites can be found?
I have observed that, only one Cipher Suite is specified which is : TLS_EMPTY_RENEGOTIATION_INFO_SCSV
This is no real cipher. If no other ciphers are specified then the client does not offer any ciphers at all which means that no shared ciphers can be found and thus the handshake will fail. It looks like the client is buggy. Reason might be a failed attempt to fight POODLE attack by disabling all SSL3.0 ciphers, which in effect disables all ciphers for TLS1 1.0 and TLS 1.1.
Following a really outdated tutorial I managed to create an HTTPS server using OpenSSL with TLS1.2, and I'm very proud of it ;)
However TLS 1.2 is only supported in latest browsers and I would like to have some kind of negotiation of the protocol between the client and server, which I'm sure it can be done, but I'm not able to find how! So that if the client only supports TLS1.0, well use that. And if it only supports SSLv3, use that. Not sure about SSLv2, maybe better leave that...
The code I use right now is:
SSL_library_init();
OpenSSL_add_all_algorithms();
SSL_load_error_strings();
ssl_method = TLSv1_2_server_method();
ssl_ctx = SSL_CTX_new(ssl_method);
Then the server certificates are loaded and the ssl_ctx is shared among all connections. When a client is accepted by the server socket it is encapsulated in an SSL object (whatever it represents):
ssl = SSL_new(ssl_ctx);
SSL_set_fd(ssl, client_socket);
SSL_accept(ssl);
So I guess that something has to be changed in the ssl_ctx creation to allow more methods... Any idea?
<rant> No decent, extensive documentation can be found for OpenSSL, the best available is a 10 years old tutorial! </rant>
Thanks in advance.
You do this by using SSLv23_method() (and friends) instead of a specific method (e.g. TLSv1_2_server_method() in your example). This sends the SSLv2 ClientHello but also specifies the highest protocol supported. The somewhat outdated man page says:
SSLv23_method(void), SSLv23_server_method(void), SSLv23_client_method(void)
A TLS/SSL connection established with these methods will understand
the SSLv2, SSLv3, and TLSv1 protocol. A client will send out SSLv2
client hello messages and will indicate that it also understands SSLv3
and TLSv1. A server will understand SSLv2, SSLv3, and TLSv1 client
hello messages. This is the best choice when compatibility is a
concern.
This online man page doesn't discuss the newer TLSv1_1 and TLSv1_2 protocols, but I verified in the 1.0.1g source of s23_clnt.c that SSLv23_method() includes them.
You then limit the protocols you actually accept with SSL_CTX_set_options():
The list of protocols available can later be limited using the
SSL_OP_NO_SSLv2, SSL_OP_NO_SSLv3, SSL_OP_NO_TLSv1 options of the
SSL_CTX_set_options() or SSL_set_options() functions. Using these
options it is possible to choose e.g. SSLv23_server_method() and be
able to negotiate with all possible clients, but to only allow newer
protocols like SSLv3 or TLSv1.
Note, however, that you can't enable arbitrary sets of protocols, only contiguous protocols in SSLv2, SSLv3, TLSv1, TLSv1_1, TLSv1_2. For example, you can't choose only SSLv3 and TLSv1_1, omitting TLSv1. This comment in the source explains why:
SSL_OP_NO_X disables all protocols above X if there are some protocols below X enabled. This is required in order to maintain "version capability" vector contiguous. So that if application wants to disable TLS1.0 in favour of TLS1>=1, it would be insufficient to pass SSL_NO_TLSv1, the answer is SSL_OP_NO_TLSv1|SSL_OP_NO_SSLv3|SSL_OP_NO_SSLv2.