--- gvpe/doc/gvpe.protocol.7 2005/03/17 22:24:31 1.3 +++ gvpe/doc/gvpe.protocol.7 2013/09/20 11:57:03 1.13 @@ -1,15 +1,7 @@ -.\" Automatically generated by Pod::Man v1.37, Pod::Parser v1.14 +.\" Automatically generated by Pod::Man 2.25 (Pod::Simple 3.20) .\" .\" Standard preamble: .\" ======================================================================== -.de Sh \" Subsection heading -.br -.if t .Sp -.ne 5 -.PP -\fB\\$1\fR -.PP -.. .de Sp \" Vertical space (when we can't use .PP) .if t .sp .5v .if n .sp @@ -25,11 +17,11 @@ .. .\" Set up some character translations and predefined strings. \*(-- will .\" give an unbreakable dash, \*(PI will give pi, \*(L" will give a left -.\" double quote, and \*(R" will give a right double quote. | will give a -.\" real vertical bar. \*(C+ will give a nicer C++. Capital omega is used to -.\" do unbreakable dashes and therefore won't be available. \*(C` and \*(C' -.\" expand to `' in nroff, nothing in troff, for use with C<>. -.tr \(*W-|\(bv\*(Tr +.\" double quote, and \*(R" will give a right double quote. \*(C+ will +.\" give a nicer C++. Capital omega is used to do unbreakable dashes and +.\" therefore won't be available. \*(C` and \*(C' expand to `' in nroff, +.\" nothing in troff, for use with C<>. +.tr \(*W- .ds C+ C\v'-.1v'\h'-1p'\s-2+\h'-1p'+\s0\v'.1v'\h'-1p' .ie n \{\ . ds -- \(*W- @@ -48,22 +40,25 @@ . ds R" '' 'br\} .\" +.\" Escape single quotes in literal strings from groff's Unicode transform. +.ie \n(.g .ds Aq \(aq +.el .ds Aq ' +.\" .\" If the F register is turned on, we'll generate index entries on stderr for -.\" titles (.TH), headers (.SH), subsections (.Sh), items (.Ip), and index +.\" titles (.TH), headers (.SH), subsections (.SS), items (.Ip), and index .\" entries marked with X<> in POD. Of course, you'll have to process the .\" output yourself in some meaningful fashion. -.if \nF \{\ +.ie \nF \{\ . de IX . tm Index:\\$1\t\\n%\t"\\$2" .. . nr % 0 . rr F .\} -.\" -.\" For nroff, turn off justification. Always turn off hyphenation; it makes -.\" way too many mistakes in technical documents. -.hy 0 -.if n .na +.el \{\ +. de IX +.. +.\} .\" .\" Accent mark definitions (@(#)ms.acc 1.5 88/02/08 SMI; from UCB 4.2). .\" Fear. Run. Save yourself. No user-serviceable parts. @@ -129,7 +124,11 @@ .\" ======================================================================== .\" .IX Title "GVPE.PROTOCOL 7" -.TH GVPE.PROTOCOL 7 "2005-03-17" "1.8" "GNU Virtual Private Ethernet" +.TH GVPE.PROTOCOL 7 "2013-07-19" "2.25" "GNU Virtual Private Ethernet" +.\" For nroff, turn off justification. Always turn off hyphenation; it makes +.\" way too many mistakes in technical documents. +.if n .ad l +.nh .SH "The GNU-VPE Protocols" .IX Header "The GNU-VPE Protocols" .SH "Overview" @@ -141,46 +140,46 @@ .PP The first part of this document describes the transport protocols which are used by \s-1GVPE\s0 to send it's data packets over the network. -.SH "PART 1: Tansport protocols" -.IX Header "PART 1: Tansport protocols" -\&\s-1GVPE\s0 offers a range of transport protocols that can be used to interchange -data between nodes. Protocols differ in their overhead, speed, +.SH "PART 1: Transport protocols" +.IX Header "PART 1: Transport protocols" +\&\s-1GVPE\s0 offers a wide range of transport protocols that can be used to +interchange data between nodes. Protocols differ in their overhead, speed, reliability, and robustness. .PP The following sections describe each transport protocol in more detail. They are sorted by overhead/efficiency, the most efficient -transprot is listed first: -.Sh "\s-1RAW\s0 \s-1IP\s0" +transport is listed first: +.SS "\s-1RAW\s0 \s-1IP\s0" .IX Subsection "RAW IP" -This protocol is the best choice, performance\-wise, as the minimum +This protocol is the best choice, performance-wise, as the minimum overhead per packet is only 38 bytes. .PP -It works by sending the \s-1VPN\s0 payload using raw ip frames (using the +It works by sending the \s-1VPN\s0 payload using raw \s-1IP\s0 frames (using the protocol set by \f(CW\*(C`ip\-proto\*(C'\fR). .PP -Using raw ip frames has the drawback that many firewalls block \*(L"unknown\*(R" +Using raw \s-1IP\s0 frames has the drawback that many firewalls block \*(L"unknown\*(R" protocols, so this transport only works if you have full \s-1IP\s0 connectivity between nodes. -.Sh "\s-1ICMP\s0" +.SS "\s-1ICMP\s0" .IX Subsection "ICMP" This protocol offers very low overhead (minimum 42 bytes), and can -sometimes tunnel through firewalls when other protocols cannot. +sometimes tunnel through firewalls when other protocols can not. .PP -It works by prepending a \s-1ICMP\s0 header with type \f(CW\*(C`icmp\-type\*(C'\fR and a code +It works by prepending an \s-1ICMP\s0 header with type \f(CW\*(C`icmp\-type\*(C'\fR and a code of \f(CW255\fR. The default \f(CW\*(C`icmp\-type\*(C'\fR is \f(CW\*(C`echo\-reply\*(C'\fR, so the resulting packets look like echo replies, which looks rather strange to network -admins. +administrators. .PP -This transport should only be used if other transports (i.e. raw ip) are +This transport should only be used if other transports (i.e. raw \s-1IP\s0) are not available or undesirable (due to their overhead). -.Sh "\s-1UDP\s0" +.SS "\s-1UDP\s0" .IX Subsection "UDP" This is a good general choice for the transport protocol as \s-1UDP\s0 packets tunnel well through most firewalls and routers, and the overhead per packet is moderate (minimum 58 bytes). .PP It should be used if \s-1RAW\s0 \s-1IP\s0 is not available. -.Sh "\s-1TCP\s0" +.SS "\s-1TCP\s0" .IX Subsection "TCP" This protocol is a very bad choice, as it not only has high overhead (more than 60 bytes), but the transport also retries on it's own, which leads @@ -197,11 +196,11 @@ It is an abuse of the usage a proxy was designed for, so make sure you are allowed to use it for \s-1GVPE\s0. .PP -This protocol also has server and client sides. If the \f(CW\*(C`tcp\-port\*(C'\fR is set -to zero, other nodes cannot connect to this node directly (and \f(CW\*(C`tcp\-port\*(C'\fR -zero cannot be used). If the \f(CW\*(C`tcp\-port\*(C'\fR is non\-zero, the node can act -both as a client as well as a server. -.Sh "\s-1DNS\s0" +This protocol also has server and client sides. If the \f(CW\*(C`tcp\-port\*(C'\fR is +set to zero, other nodes cannot connect to this node directly. If the +\&\f(CW\*(C`tcp\-port\*(C'\fR is non-zero, the node can act both as a client as well as a +server. +.SS "\s-1DNS\s0" .IX Subsection "DNS" \&\fB\s-1WARNING:\s0\fR Parsing and generating \s-1DNS\s0 packets is rather tricky. The code almost certainly contains buffer overflows and other, likely exploitable, @@ -217,34 +216,34 @@ In addition, the same problems as the \s-1TCP\s0 transport also plague this protocol. .PP -Most configuration needs to be done by editing \f(CW\*(C`src/vpn_dns.C\*(C'\fR directly. -.PP -It's only use is to tunnel through firewalls that do not allow direct +Its only use is to tunnel through firewalls that do not allow direct internet access. Similar to using a \s-1HTTP\s0 proxy (as the \s-1TCP\s0 transport does), it uses a local \s-1DNS\s0 server/forwarder (given by the \f(CW\*(C`dns\-forw\-host\*(C'\fR configuration value) as a proxy to send and receive data as a client, -and a \f(CW\*(C`NS\*(C'\fR record pointing to the \s-1GVPE\s0 server (as given by the +and an \f(CW\*(C`NS\*(C'\fR record pointing to the \s-1GVPE\s0 server (as given by the \&\f(CW\*(C`dns\-hostname\*(C'\fR directive). .PP The only good side of this protocol is that it can tunnel through most -firewalls undetected, iff the local \s-1DNS\s0 server/forwarder is sane (which is -true for most routers, wlan gateways and nameservers). +firewalls mostly undetected, iff the local \s-1DNS\s0 server/forwarder is sane +(which is true for most routers, wireless \s-1LAN\s0 gateways and nameservers). +.PP +Fine-tuning needs to be done by editing \f(CW\*(C`src/vpn_dns.C\*(C'\fR directly. .SH "PART 2: The GNU VPE protocol" .IX Header "PART 2: The GNU VPE protocol" This section, unfortunately, is not yet finished, although the protocol is stable (until bugs in the cryptography are found, which will likely completely change the following description). Nevertheless, it should give you some overview over the protocol. -.Sh "Anatomy of a \s-1VPN\s0 packet" +.SS "Anatomy of a \s-1VPN\s0 packet" .IX Subsection "Anatomy of a VPN packet" The exact layout and field lengths of a \s-1VPN\s0 packet is determined at -compiletime and doesn't change. The same structure is used for all -transort protocols, be it \s-1RAWIP\s0 or \s-1TCP\s0. +compile time and doesn't change. The same structure is used for all +transport protocols, be it \s-1RAWIP\s0 or \s-1TCP\s0. .PP .Vb 3 -\& +------+------+--------+------+ +\& +\-\-\-\-\-\-+\-\-\-\-\-\-+\-\-\-\-\-\-\-\-+\-\-\-\-\-\-+ \& | HMAC | TYPE | SRCDST | DATA | -\& +------+------+--------+------+ +\& +\-\-\-\-\-\-+\-\-\-\-\-\-+\-\-\-\-\-\-\-\-+\-\-\-\-\-\-+ .Ve .PP The \s-1HMAC\s0 field is present in all packets, even if not used (e.g. in auth @@ -256,102 +255,160 @@ \&\s-1CONNECT\s0 \s-1REQUEST/INFO\s0 etc.). .PP \&\s-1SRCDST\s0 is a three byte field which contains the source and destination -node ids (12 bits each). The protocol does not yet scale well beyond 30+ -hosts, since all hosts must connect to each other once on startup. But if -restarts are rare or tolerable and most connections are on demand, much -larger networks are feasible. +node IDs (12 bits each). .PP The \s-1DATA\s0 portion differs between each packet type, naturally, and is the only part that can be encrypted. Data packets contain more fields, as shown: .PP .Vb 3 -\& +------+------+--------+------+-------+------+ +\& +\-\-\-\-\-\-+\-\-\-\-\-\-+\-\-\-\-\-\-\-\-+\-\-\-\-\-\-+\-\-\-\-\-\-\-+\-\-\-\-\-\-+ \& | HMAC | TYPE | SRCDST | RAND | SEQNO | DATA | -\& +------+------+--------+------+-------+------+ +\& +\-\-\-\-\-\-+\-\-\-\-\-\-+\-\-\-\-\-\-\-\-+\-\-\-\-\-\-+\-\-\-\-\-\-\-+\-\-\-\-\-\-+ .Ve .PP \&\s-1RAND\s0 is a sequence of fully random bytes, used to increase the entropy of the data for encryption purposes. .PP \&\s-1SEQNO\s0 is a 32\-bit sequence number. It is negotiated at every connection -initialization and starts at some random 31 bit value. \s-1VPE\s0 currently uses +initialization and starts at some random 31 bit value. \s-1GVPE\s0 currently uses a sliding window of 512 packets/sequence numbers to detect reordering, -duplication and reply attacks. -.Sh "The authentification protocol" -.IX Subsection "The authentification protocol" -Before hosts can exchange packets, they need to establish authenticity of -the other side and a key. Every host has a private \s-1RSA\s0 key and the public -\&\s-1RSA\s0 keys of all other hosts. -.PP -A host establishes a simplex connection by sending the other host a -\&\s-1RSA\s0 encrypted challenge containing a random challenge (consisting of -the encryption key to use when sending packets, more random data and -\&\s-1PKCS1_OAEP\s0 padding) and a random 16 byte \*(L"challenge\-id\*(R" (used to detect -duplicate auth packets). The destination host will respond by replying -with an (unencrypted) \s-1RIPEMD160\s0 hash of the decrypted challenge, which -will authentify that host. The destination host will also set the outgoing -encryption parameters as given in the packet. -.PP -When the source host receives a correct auth reply (by verifying the -hash and the id, which will expire after 120 seconds), it will start to -accept data packets from the destination host. -.PP -This means that a host can only initate a simplex connection, telling the -other side the key it has to use when it sends packets. The challenge -reply is only used to set the current \s-1IP\s0 address of the other side and -protocol parameters. -.PP -This protocol is completely symmetric, so to be able to send packets the -destination host must send a challenge in the exact same way as already -described (so, in essence, two simplex connections are created per host -pair). -.Sh "Retrying" +duplication and replay attacks. +.PP +The encryption is done on \s-1RAND+SEQNO+DATA\s0 in \s-1CBC\s0 mode with zero \s-1IV\s0 (or, +equivalently, the \s-1IV\s0 is \s-1RAND+SEQNO\s0, encrypted with the block cipher, +unless \s-1RAND\s0 size is decreased or increased over the default value). +.PP +The random prefix itself is generated by using \s-1AES\s0 in \s-1CTR\s0 mode with a +random key and starting value, which should make them unpredictable even +before encrypting them again. The sequence number additionally ensures +that the \s-1IV\s0 is unique. +.SS "The authentication/key exchange protocol" +.IX Subsection "The authentication/key exchange protocol" +Before nodes can exchange packets, they need to establish authenticity of +the other side and a key. Every node has a private \s-1RSA\s0 key and the public +\&\s-1RSA\s0 keys of all other nodes. +.PP +When a node wants to establish a connection to another node, it sends an +RSA-OEAP-encrypted challenge and an \s-1ECDH\s0 key. The other node replies with +it's own \s-1ECDH\s0 key and a \s-1HKDF\s0 of the challange and both \s-1ECDH\s0 keys to proof +it's identity. +.PP +The remote node enganges in exactly the same protocol. When both nodes +have exchanged their challenge and verified the response, they calculate a +cipher key and a \s-1HMAC\s0 key and start exchanging data packets. +.PP +In detail, the challenge consist of: +.PP +.Vb 1 +\& RSA\-OAEP (SEQNO MAC CIPHER SALT EXTRA\-AUTH) ECDH1 +.Ve +.PP +That is, it encrypts (with the public key of the remote node) an initial +sequence number for data packets, key material for the \s-1HMAC\s0 key, key +material for the cipher key, a salt used by the \s-1HKDF\s0 (as shown later) and +some extra random bytes that are unused except for authentication. It also +sends the public key of a curve25519 exchange. +.PP +The remote node decrypts the \s-1RSA\s0 data, generates it's own \s-1ECDH\s0 key (\s-1ECDH2\s0), and +replies with: +.PP +.Vb 1 +\& HKDF\-Expand (HKDF\-Extract (ECDH2, RSA), ECDH1, AUTH_DIGEST_SIZE) ECDH2 +.Ve +.PP +That is, it extracts from the decrypted \s-1RSA\s0 challenge, using it's \s-1ECDH\s0 +key as salt, and then expands using the requesting node's \s-1ECDH1\s0 key. The +resulting has is returned as a proof that the node could decrypt the \s-1RSA\s0 +challenge data, together with the \s-1ECDH\s0 key. +.PP +After both nodes have done this to each other, they calculate the shared +\&\s-1ECDH\s0 secrets, cipher and \s-1HMAC\s0 keys for the session (each +node generates two cipher and \s-1HMAC\s0 keys, one for sending and one for +receiving). +.PP +The \s-1HMAC\s0 key for sending is generated as follow: +.PP +.Vb 1 +\& HMAC_KEY = HKDF\-Expand (HKDF\-Extract (REMOTE_SALT, MAC ECDH_SECRET), info, HMAC_MD_SIZE) +.Ve +.PP +It extracts from \s-1MAC\s0 and \s-1ECDH_SECRET\s0 using the \fIremote\fR \s-1SALT\s0, then +expands using a static info string. +.PP +The cipher key is generated in the same way, except using the \s-1CIPHER\s0 part +of the original challenge. +.PP +The result of this process is to authenticate each node to the other +node, while exchanging keys using both \s-1RSA\s0 and \s-1ECDH\s0, the latter providing +perfect forward secrecy. +.PP +The protocol has been overdesigned where this was possible without +increasing implementation complexity, in an attempt to protect against +implementation or protocol failures. For example, if the \s-1ECDH\s0 challenge +was found to be flawed, perfect forward secrecy would be lost, but +the data would still be protected. Likewise, standard algorithms and +implementations are used where possible. +.SS "Retrying" .IX Subsection "Retrying" -When there is no response to an auth request, the host will send auth -requests in bursts with an exponential backoff. After some time it will -resort to \s-1PING\s0 packets, which are very small (8 bytes) and lightweight -(no \s-1RSA\s0 operations required). A host that receives ping requests from an -unconnected peer will respond by trying to create a connection. +When there is no response to an auth request, the node will send auth +requests in bursts with an exponential back-off. After some time it will +resort to \s-1PING\s0 packets, which are very small (8 bytes + protocol header) +and lightweight (no \s-1RSA\s0 operations required). A node that receives ping +requests from an unconnected peer will respond by trying to create a +connection. .PP -In addition to the exponential backoff, there is a global rate-limit on +In addition to the exponential back-off, there is a global rate-limit on a per-IP base. It allows long bursts but will limit total packet rate to something like one control packet every ten seconds, to avoid accidental floods due to protocol problems (like a \s-1RSA\s0 key file mismatch between two -hosts). -.Sh "Routing and Protocol translation" +nodes). +.PP +The intervals between retries are limited by the \f(CW\*(C`max\-retry\*(C'\fR +configuration value. A node with \f(CW\*(C`connect\*(C'\fR = \f(CW\*(C`always\*(C'\fR will always retry, +a node with \f(CW\*(C`connect\*(C'\fR = \f(CW\*(C`ondemand\*(C'\fR will only try (and re-try) to connect +as long as there are packets in the queue, usually this limits the retry +period to \f(CW\*(C`max\-ttl\*(C'\fR seconds. +.PP +Sending packets over the \s-1VPN\s0 will reset the retry intervals as well, which +means as long as somebody is trying to send packets to a given node, \s-1GVPE\s0 +will try to connect every few seconds. +.SS "Routing and Protocol translation" .IX Subsection "Routing and Protocol translation" -The gvpe routing algorithm is easy: there isn't any routing. \s-1GVPE\s0 always -tries to establish direct connections, if the protocol abilities of the -two hosts allow it. -.PP -If the two hosts should be able to reach each other (common protocol, ip -and port all known), but cannot (network down), then there will be no -connection, point. +The \s-1GVPE\s0 routing algorithm is easy: there isn't much routing to speak +of: When routing packets to another node, \s-1GVPE\s0 tries the following +options, in order: +.IP "If the two nodes should be able to reach each other directly (common protocol, port known), then \s-1GVPE\s0 will send the packet directly to the other node." 4 +.IX Item "If the two nodes should be able to reach each other directly (common protocol, port known), then GVPE will send the packet directly to the other node." +.PD 0 +.ie n .IP "If this isn't possible (e.g. because the node doesn't have a \*(C`hostname\*(C' or known port), but the nodes speak a common protocol and a router is available, then \s-1GVPE\s0 will ask a router to ""mediate"" between both nodes (see below)." 4 +.el .IP "If this isn't possible (e.g. because the node doesn't have a \f(CW\*(C`hostname\*(C'\fR or known port), but the nodes speak a common protocol and a router is available, then \s-1GVPE\s0 will ask a router to ``mediate'' between both nodes (see below)." 4 +.IX Item "If this isn't possible (e.g. because the node doesn't have a hostname or known port), but the nodes speak a common protocol and a router is available, then GVPE will ask a router to mediate between both nodes (see below)." +.ie n .IP "If a direct connection isn't possible (no common protocols) or forbidden (\*(C`deny\-direct\*(C') and there are any routers, then \s-1GVPE\s0 will try to send packets to the router with the highest priority that is connected already \fIand\fR is able (as specified by the config file) to connect directly to the target node." 4 +.el .IP "If a direct connection isn't possible (no common protocols) or forbidden (\f(CW\*(C`deny\-direct\*(C'\fR) and there are any routers, then \s-1GVPE\s0 will try to send packets to the router with the highest priority that is connected already \fIand\fR is able (as specified by the config file) to connect directly to the target node." 4 +.IX Item "If a direct connection isn't possible (no common protocols) or forbidden (deny-direct) and there are any routers, then GVPE will try to send packets to the router with the highest priority that is connected already and is able (as specified by the config file) to connect directly to the target node." +.IP "If no such router exists, then \s-1GVPE\s0 will simply send the packet to the node with the highest priority available." 4 +.IX Item "If no such router exists, then GVPE will simply send the packet to the node with the highest priority available." +.IP "Failing all that, the packet will be dropped." 4 +.IX Item "Failing all that, the packet will be dropped." +.PD .PP A host can usually declare itself unreachable directly by setting it's port number(s) to zero. It can declare other hosts as unreachable by using -a config-file that disables all protocols for these other hosts. +a config-file that disables all protocols for these other hosts. Another +option is to disable all protocols on that host in the other config files. .PP If two hosts cannot connect to each other because their \s-1IP\s0 address(es) -are not known (such as dialup hosts), one side will send a connection -request to a router (routers must be configured to act as routers!), which -will send both the originating and the destination host a connection info -request with protocol information and \s-1IP\s0 address of the other host (if -known). Both hosts will then try to establish a connection to the other -peer, which is usually possible even when both hosts are behind a \s-1NAT\s0 -gateway. -.PP -If the hosts cannot reach each other because they have no common protocol, -the originator instead use the router with highest priority and matching -protocol as peer. Since the \s-1SRCDST\s0 field is not encrypted, the router host -can just forward the packet to the destination host. Since each host uses -it's own private key, the router will not be able to decrypt or encrypt -packets, it will just act as a simple router and protocol translator. -.PP -When no router is connected, the host will aggressively try to connect to -all routers, and if a router is asked for an unconnected host it will try -to ask another router to establish the connection. -.PP -\&... more not yet written about the details of the routing, please bug me -\&... +are not known (such as dial-up hosts), one side will send a \fImediated\fR +connection request to a router (routers must be configured to act as +routers!), which will send both the originating and the destination host +a connection info request with protocol information and \s-1IP\s0 address of the +other host (if known). Both hosts will then try to establish a direct +connection to the other peer, which is usually possible even when both +hosts are behind a \s-1NAT\s0 gateway. +.PP +Routing via other nodes works because the \s-1SRCDST\s0 field is not encrypted, +so the router can just forward the packet to the destination host. Since +each host uses it's own private key, the router will not be able to +decrypt or encrypt packets, it will just act as a simple router and +protocol translator.