You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
772 lines
22 KiB
772 lines
22 KiB
/*
|
|
This file is part of cpp-ethereum.
|
|
|
|
cpp-ethereum is free software: you can redistribute it and/or modify
|
|
it under the terms of the GNU General Public License as published by
|
|
the Free Software Foundation, either version 3 of the License, or
|
|
(at your option) any later version.
|
|
|
|
cpp-ethereum is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
GNU General Public License for more details.
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
along with cpp-ethereum. If not, see <http://www.gnu.org/licenses/>.
|
|
*/
|
|
/** @file Host.cpp
|
|
* @author Alex Leverington <nessence@gmail.com>
|
|
* @author Gav Wood <i@gavwood.com>
|
|
* @date 2014
|
|
*/
|
|
|
|
#include <set>
|
|
#include <chrono>
|
|
#include <thread>
|
|
#include <mutex>
|
|
#include <memory>
|
|
#include <boost/algorithm/string.hpp>
|
|
#include <libdevcore/Common.h>
|
|
#include <libdevcore/CommonIO.h>
|
|
#include <libdevcore/StructuredLogger.h>
|
|
#include <libethcore/Exceptions.h>
|
|
#include <libdevcrypto/FileSystem.h>
|
|
#include "Session.h"
|
|
#include "Common.h"
|
|
#include "Capability.h"
|
|
#include "UPnP.h"
|
|
#include "RLPxHandshake.h"
|
|
#include "Host.h"
|
|
using namespace std;
|
|
using namespace dev;
|
|
using namespace dev::p2p;
|
|
|
|
HostNodeTableHandler::HostNodeTableHandler(Host& _host): m_host(_host) {}
|
|
|
|
void HostNodeTableHandler::processEvent(NodeId const& _n, NodeTableEventType const& _e)
|
|
{
|
|
m_host.onNodeTableEvent(_n, _e);
|
|
}
|
|
|
|
Host::Host(std::string const& _clientVersion, NetworkPreferences const& _n, bytesConstRef _restoreNetwork):
|
|
Worker("p2p", 0),
|
|
m_restoreNetwork(_restoreNetwork.toBytes()),
|
|
m_clientVersion(_clientVersion),
|
|
m_netPrefs(_n),
|
|
m_ifAddresses(Network::getInterfaceAddresses()),
|
|
m_ioService(2),
|
|
m_tcp4Acceptor(m_ioService),
|
|
m_alias(networkAlias(_restoreNetwork)),
|
|
m_lastPing(chrono::steady_clock::time_point::min())
|
|
{
|
|
for (auto address: m_ifAddresses)
|
|
if (address.is_v4())
|
|
clog(NetNote) << "IP Address: " << address << " = " << (isPrivateAddress(address) ? "[LOCAL]" : "[PEER]");
|
|
|
|
clog(NetNote) << "Id:" << id();
|
|
}
|
|
|
|
Host::~Host()
|
|
{
|
|
stop();
|
|
}
|
|
|
|
void Host::start()
|
|
{
|
|
startWorking();
|
|
}
|
|
|
|
void Host::stop()
|
|
{
|
|
// called to force io_service to kill any remaining tasks it might have -
|
|
// such tasks may involve socket reads from Capabilities that maintain references
|
|
// to resources we're about to free.
|
|
|
|
{
|
|
// Although m_run is set by stop() or start(), it effects m_runTimer so x_runTimer is used instead of a mutex for m_run.
|
|
// when m_run == false, run() will cause this::run() to stop() ioservice
|
|
Guard l(x_runTimer);
|
|
// ignore if already stopped/stopping
|
|
if (!m_run)
|
|
return;
|
|
m_run = false;
|
|
}
|
|
|
|
// wait for m_timer to reset (indicating network scheduler has stopped)
|
|
while (!!m_timer)
|
|
this_thread::sleep_for(chrono::milliseconds(50));
|
|
|
|
// stop worker thread
|
|
stopWorking();
|
|
}
|
|
|
|
void Host::doneWorking()
|
|
{
|
|
// reset ioservice (allows manually polling network, below)
|
|
m_ioService.reset();
|
|
|
|
// shutdown acceptor
|
|
m_tcp4Acceptor.cancel();
|
|
if (m_tcp4Acceptor.is_open())
|
|
m_tcp4Acceptor.close();
|
|
|
|
// There maybe an incoming connection which started but hasn't finished.
|
|
// Wait for acceptor to end itself instead of assuming it's complete.
|
|
// This helps ensure a peer isn't stopped at the same time it's starting
|
|
// and that socket for pending connection is closed.
|
|
while (m_accepting)
|
|
m_ioService.poll();
|
|
|
|
// stop capabilities (eth: stops syncing or block/tx broadcast)
|
|
for (auto const& h: m_capabilities)
|
|
h.second->onStopping();
|
|
|
|
// disconnect pending handshake, before peers, as a handshake may create a peer
|
|
for (unsigned n = 0;; n = 0)
|
|
{
|
|
{
|
|
Guard l(x_connecting);
|
|
for (auto i: m_connecting)
|
|
if (auto h = i.lock())
|
|
{
|
|
h->cancel();
|
|
n++;
|
|
}
|
|
}
|
|
if (!n)
|
|
break;
|
|
m_ioService.poll();
|
|
}
|
|
|
|
// disconnect peers
|
|
for (unsigned n = 0;; n = 0)
|
|
{
|
|
{
|
|
RecursiveGuard l(x_sessions);
|
|
for (auto i: m_sessions)
|
|
if (auto p = i.second.lock())
|
|
if (p->isConnected())
|
|
{
|
|
p->disconnect(ClientQuit);
|
|
n++;
|
|
}
|
|
}
|
|
if (!n)
|
|
break;
|
|
|
|
// poll so that peers send out disconnect packets
|
|
m_ioService.poll();
|
|
}
|
|
|
|
// stop network (again; helpful to call before subsequent reset())
|
|
m_ioService.stop();
|
|
|
|
// reset network (allows reusing ioservice in future)
|
|
m_ioService.reset();
|
|
|
|
// finally, clear out peers (in case they're lingering)
|
|
RecursiveGuard l(x_sessions);
|
|
m_sessions.clear();
|
|
}
|
|
|
|
unsigned Host::protocolVersion() const
|
|
{
|
|
return 3;
|
|
}
|
|
|
|
void Host::startPeerSession(Public const& _id, RLP const& _rlp, RLPXFrameIO* _io, bi::tcp::endpoint _endpoint)
|
|
{
|
|
/// Get or create Peer
|
|
shared_ptr<Peer> p;
|
|
p = m_peers[_id];
|
|
if (!p)
|
|
{
|
|
p.reset(new Peer()); // this maybe redundant
|
|
p->id = _id;
|
|
}
|
|
p->m_lastDisconnect = NoDisconnect;
|
|
if (p->isOffline())
|
|
p->m_lastConnected = std::chrono::system_clock::now();
|
|
p->m_failedAttempts = 0;
|
|
p->endpoint.tcp.address(_endpoint.address());
|
|
|
|
auto protocolVersion = _rlp[0].toInt<unsigned>();
|
|
auto clientVersion = _rlp[1].toString();
|
|
auto caps = _rlp[2].toVector<CapDesc>();
|
|
auto listenPort = _rlp[3].toInt<unsigned short>();
|
|
|
|
// clang error (previously: ... << hex << caps ...)
|
|
// "'operator<<' should be declared prior to the call site or in an associated namespace of one of its arguments"
|
|
stringstream capslog;
|
|
for (auto cap: caps)
|
|
capslog << "(" << cap.first << "," << dec << cap.second << ")";
|
|
clog(NetMessageSummary) << "Hello: " << clientVersion << "V[" << protocolVersion << "]" << _id.abridged() << showbase << capslog.str() << dec << listenPort;
|
|
|
|
// create session so disconnects are managed
|
|
auto ps = make_shared<Session>(this, _io, p, PeerSessionInfo({_id, clientVersion, _endpoint.address().to_string(), listenPort, chrono::steady_clock::duration(), _rlp[2].toSet<CapDesc>(), 0, map<string, string>()}));
|
|
if (protocolVersion != this->protocolVersion())
|
|
{
|
|
ps->disconnect(IncompatibleProtocol);
|
|
return;
|
|
}
|
|
|
|
{
|
|
RecursiveGuard l(x_sessions);
|
|
if (m_sessions.count(_id) && !!m_sessions[_id].lock())
|
|
if (auto s = m_sessions[_id].lock())
|
|
if(s->isConnected())
|
|
{
|
|
// Already connected.
|
|
clog(NetWarn) << "Session already exists for peer with id" << _id.abridged();
|
|
ps->disconnect(DuplicatePeer);
|
|
return;
|
|
}
|
|
m_sessions[_id] = ps;
|
|
}
|
|
ps->start();
|
|
unsigned o = (unsigned)UserPacket;
|
|
for (auto const& i: caps)
|
|
if (haveCapability(i))
|
|
{
|
|
ps->m_capabilities[i] = shared_ptr<Capability>(m_capabilities[i]->newPeerCapability(ps.get(), o));
|
|
o += m_capabilities[i]->messageCount();
|
|
}
|
|
clog(NetNote) << "p2p.host.peer.register" << _id.abridged();
|
|
StructuredLogger::p2pConnected(_id.abridged(), ps->m_peer->peerEndpoint(), ps->m_peer->m_lastConnected, clientVersion, peerCount());
|
|
}
|
|
|
|
void Host::onNodeTableEvent(NodeId const& _n, NodeTableEventType const& _e)
|
|
{
|
|
|
|
if (_e == NodeEntryAdded)
|
|
{
|
|
clog(NetNote) << "p2p.host.nodeTable.events.nodeEntryAdded " << _n;
|
|
|
|
auto n = m_nodeTable->node(_n);
|
|
if (n)
|
|
{
|
|
shared_ptr<Peer> p;
|
|
{
|
|
RecursiveGuard l(x_sessions);
|
|
if (m_peers.count(_n))
|
|
p = m_peers[_n];
|
|
else
|
|
{
|
|
// TODO p2p: construct peer from node
|
|
p.reset(new Peer());
|
|
p->id = _n;
|
|
p->endpoint = NodeIPEndpoint(n.endpoint.udp, n.endpoint.tcp);
|
|
p->required = n.required;
|
|
m_peers[_n] = p;
|
|
|
|
clog(NetNote) << "p2p.host.peers.events.peersAdded " << _n << p->endpoint.tcp.address() << p->endpoint.udp.address();
|
|
}
|
|
p->endpoint.tcp = n.endpoint.tcp;
|
|
}
|
|
|
|
// TODO: Implement similar to discover. Attempt connecting to nodes
|
|
// until ideal peer count is reached; if all nodes are tried,
|
|
// repeat. Notably, this is an integrated process such that
|
|
// when onNodeTableEvent occurs we should also update +/-
|
|
// the list of nodes to be tried. Thus:
|
|
// 1) externalize connection attempts
|
|
// 2) attempt copies potentialPeer list
|
|
// 3) replace this logic w/maintenance of potentialPeers
|
|
if (peerCount() < m_idealPeerCount)
|
|
connect(p);
|
|
}
|
|
}
|
|
else if (_e == NodeEntryRemoved)
|
|
{
|
|
clog(NetNote) << "p2p.host.nodeTable.events.nodeEntryRemoved " << _n;
|
|
|
|
RecursiveGuard l(x_sessions);
|
|
m_peers.erase(_n);
|
|
}
|
|
}
|
|
|
|
void Host::determinePublic(string const& _publicAddress, bool _upnp)
|
|
{
|
|
m_peerAddresses.clear();
|
|
|
|
// no point continuing if there are no interface addresses or valid listen port
|
|
if (!m_ifAddresses.size() || m_listenPort < 1)
|
|
return;
|
|
|
|
// populate interfaces we'll listen on (eth listens on all interfaces); ignores local
|
|
for (auto addr: m_ifAddresses)
|
|
if ((m_netPrefs.localNetworking || !isPrivateAddress(addr)) && !isLocalHostAddress(addr))
|
|
m_peerAddresses.insert(addr);
|
|
|
|
// if user supplied address is a public address then we use it
|
|
// if user supplied address is private, and localnetworking is enabled, we use it
|
|
bi::address reqPublicAddr(bi::address(_publicAddress.empty() ? bi::address() : bi::address::from_string(_publicAddress)));
|
|
bi::tcp::endpoint reqPublic(reqPublicAddr, m_listenPort);
|
|
bool isprivate = isPrivateAddress(reqPublicAddr);
|
|
bool ispublic = !isprivate && !isLocalHostAddress(reqPublicAddr);
|
|
if (!reqPublicAddr.is_unspecified() && (ispublic || (isprivate && m_netPrefs.localNetworking)))
|
|
{
|
|
if (!m_peerAddresses.count(reqPublicAddr))
|
|
m_peerAddresses.insert(reqPublicAddr);
|
|
m_tcpPublic = reqPublic;
|
|
return;
|
|
}
|
|
|
|
// if address wasn't provided, then use first public ipv4 address found
|
|
for (auto addr: m_peerAddresses)
|
|
if (addr.is_v4() && !isPrivateAddress(addr))
|
|
{
|
|
m_tcpPublic = bi::tcp::endpoint(*m_peerAddresses.begin(), m_listenPort);
|
|
return;
|
|
}
|
|
|
|
// or find address via upnp
|
|
if (_upnp)
|
|
{
|
|
bi::address upnpifaddr;
|
|
bi::tcp::endpoint upnpep = Network::traverseNAT(m_ifAddresses, m_listenPort, upnpifaddr);
|
|
if (!upnpep.address().is_unspecified() && !upnpifaddr.is_unspecified())
|
|
{
|
|
if (!m_peerAddresses.count(upnpep.address()))
|
|
m_peerAddresses.insert(upnpep.address());
|
|
m_tcpPublic = upnpep;
|
|
return;
|
|
}
|
|
}
|
|
|
|
// or if no address provided, use private ipv4 address if local networking is enabled
|
|
if (reqPublicAddr.is_unspecified())
|
|
if (m_netPrefs.localNetworking)
|
|
for (auto addr: m_peerAddresses)
|
|
if (addr.is_v4() && isPrivateAddress(addr))
|
|
{
|
|
m_tcpPublic = bi::tcp::endpoint(addr, m_listenPort);
|
|
return;
|
|
}
|
|
|
|
// otherwise address is unspecified
|
|
m_tcpPublic = bi::tcp::endpoint(bi::address(), m_listenPort);
|
|
}
|
|
|
|
void Host::runAcceptor()
|
|
{
|
|
assert(m_listenPort > 0);
|
|
|
|
if (m_run && !m_accepting)
|
|
{
|
|
clog(NetConnect) << "Listening on local port " << m_listenPort << " (public: " << m_tcpPublic << ")";
|
|
m_accepting = true;
|
|
|
|
auto socket = make_shared<RLPXSocket>(new bi::tcp::socket(m_ioService));
|
|
m_tcp4Acceptor.async_accept(socket->ref(), [=](boost::system::error_code ec)
|
|
{
|
|
// if no error code
|
|
bool success = false;
|
|
if (!ec)
|
|
{
|
|
try
|
|
{
|
|
// incoming connection; we don't yet know nodeid
|
|
auto handshake = make_shared<RLPXHandshake>(this, socket);
|
|
m_connecting.push_back(handshake);
|
|
handshake->start();
|
|
success = true;
|
|
}
|
|
catch (Exception const& _e)
|
|
{
|
|
clog(NetWarn) << "ERROR: " << diagnostic_information(_e);
|
|
}
|
|
catch (std::exception const& _e)
|
|
{
|
|
clog(NetWarn) << "ERROR: " << _e.what();
|
|
}
|
|
}
|
|
|
|
if (!success)
|
|
socket->ref().close();
|
|
|
|
m_accepting = false;
|
|
if (ec.value() < 1)
|
|
runAcceptor();
|
|
});
|
|
}
|
|
}
|
|
|
|
string Host::pocHost()
|
|
{
|
|
vector<string> strs;
|
|
boost::split(strs, dev::Version, boost::is_any_of("."));
|
|
return "poc-" + strs[1] + ".ethdev.com";
|
|
}
|
|
|
|
void Host::addNode(NodeId const& _node, std::string const& _addr, unsigned short _tcpPeerPort, unsigned short _udpNodePort)
|
|
{
|
|
// TODO: p2p clean this up (bring tested acceptor code over from network branch)
|
|
while (isWorking() && !m_run)
|
|
this_thread::sleep_for(chrono::milliseconds(50));
|
|
if (!m_run)
|
|
return;
|
|
|
|
if (_tcpPeerPort < 30300 || _tcpPeerPort > 30305)
|
|
cwarn << "Non-standard port being recorded: " << _tcpPeerPort;
|
|
|
|
if (_tcpPeerPort >= /*49152*/32768)
|
|
{
|
|
cwarn << "Private port being recorded - setting to 0";
|
|
_tcpPeerPort = 0;
|
|
}
|
|
|
|
boost::system::error_code ec;
|
|
bi::address addr = bi::address::from_string(_addr, ec);
|
|
if (ec)
|
|
{
|
|
bi::tcp::resolver *r = new bi::tcp::resolver(m_ioService);
|
|
r->async_resolve({_addr, toString(_tcpPeerPort)}, [=](boost::system::error_code const& _ec, bi::tcp::resolver::iterator _epIt)
|
|
{
|
|
if (!_ec)
|
|
{
|
|
bi::tcp::endpoint tcp = *_epIt;
|
|
if (m_nodeTable) m_nodeTable->addNode(Node(_node, NodeIPEndpoint(bi::udp::endpoint(tcp.address(), _udpNodePort), tcp)));
|
|
}
|
|
delete r;
|
|
});
|
|
}
|
|
else
|
|
if (m_nodeTable) m_nodeTable->addNode(Node(_node, NodeIPEndpoint(bi::udp::endpoint(addr, _udpNodePort), bi::tcp::endpoint(addr, _tcpPeerPort))));
|
|
}
|
|
|
|
void Host::connect(std::shared_ptr<Peer> const& _p)
|
|
{
|
|
if (!m_run)
|
|
return;
|
|
|
|
_p->m_lastAttempted = std::chrono::system_clock::now();
|
|
|
|
if (havePeerSession(_p->id))
|
|
{
|
|
clog(NetConnect) << "Aborted connect. Node already connected.";
|
|
return;
|
|
}
|
|
|
|
if (!m_nodeTable->haveNode(_p->id))
|
|
{
|
|
clog(NetWarn) << "Aborted connect. Node not in node table.";
|
|
m_nodeTable->addNode(*_p.get());
|
|
return;
|
|
}
|
|
|
|
// prevent concurrently connecting to a node
|
|
Peer *nptr = _p.get();
|
|
{
|
|
Guard l(x_pendingNodeConns);
|
|
if (m_pendingPeerConns.count(nptr))
|
|
return;
|
|
m_pendingPeerConns.insert(nptr);
|
|
}
|
|
|
|
clog(NetConnect) << "Attempting connection to node" << _p->id.abridged() << "@" << _p->peerEndpoint() << "from" << id().abridged();
|
|
auto socket = make_shared<RLPXSocket>(new bi::tcp::socket(m_ioService));
|
|
socket->ref().async_connect(_p->peerEndpoint(), [=](boost::system::error_code const& ec)
|
|
{
|
|
if (ec)
|
|
{
|
|
clog(NetConnect) << "Connection refused to node" << _p->id.abridged() << "@" << _p->peerEndpoint() << "(" << ec.message() << ")";
|
|
_p->m_lastDisconnect = TCPError;
|
|
_p->m_lastAttempted = std::chrono::system_clock::now();
|
|
}
|
|
else
|
|
{
|
|
clog(NetConnect) << "Connecting to" << _p->id.abridged() << "@" << _p->peerEndpoint();
|
|
auto handshake = make_shared<RLPXHandshake>(this, socket, _p->id);
|
|
{
|
|
Guard l(x_connecting);
|
|
m_connecting.push_back(handshake);
|
|
}
|
|
handshake->start();
|
|
}
|
|
|
|
Guard l(x_pendingNodeConns);
|
|
m_pendingPeerConns.erase(nptr);
|
|
});
|
|
}
|
|
|
|
PeerSessionInfos Host::peerSessionInfo() const
|
|
{
|
|
if (!m_run)
|
|
return PeerSessionInfos();
|
|
|
|
std::vector<PeerSessionInfo> ret;
|
|
RecursiveGuard l(x_sessions);
|
|
for (auto& i: m_sessions)
|
|
if (auto j = i.second.lock())
|
|
if (j->isConnected())
|
|
ret.push_back(j->m_info);
|
|
return ret;
|
|
}
|
|
|
|
size_t Host::peerCount() const
|
|
{
|
|
unsigned retCount = 0;
|
|
RecursiveGuard l(x_sessions);
|
|
for (auto& i: m_sessions)
|
|
if (std::shared_ptr<Session> j = i.second.lock())
|
|
if (j->isConnected())
|
|
retCount++;
|
|
return retCount;
|
|
}
|
|
|
|
void Host::run(boost::system::error_code const&)
|
|
{
|
|
if (!m_run)
|
|
{
|
|
// reset NodeTable
|
|
m_nodeTable.reset();
|
|
|
|
// stopping io service allows running manual network operations for shutdown
|
|
// and also stops blocking worker thread, allowing worker thread to exit
|
|
m_ioService.stop();
|
|
|
|
// resetting timer signals network that nothing else can be scheduled to run
|
|
m_timer.reset();
|
|
return;
|
|
}
|
|
|
|
m_nodeTable->processEvents();
|
|
|
|
// cleanup zombies
|
|
{
|
|
Guard l(x_connecting);
|
|
m_connecting.remove_if([](std::weak_ptr<RLPXHandshake> h){ return h.lock(); });
|
|
}
|
|
|
|
for (auto p: m_sessions)
|
|
if (auto pp = p.second.lock())
|
|
pp->serviceNodesRequest();
|
|
|
|
keepAlivePeers();
|
|
|
|
// At this time peers will be disconnected based on natural TCP timeout.
|
|
// disconnectLatePeers needs to be updated for the assumption that Session
|
|
// is always live and to ensure reputation and fallback timers are properly
|
|
// updated. // disconnectLatePeers();
|
|
|
|
auto openSlots = m_idealPeerCount - peerCount();
|
|
if (openSlots > 0)
|
|
{
|
|
list<shared_ptr<Peer>> toConnect;
|
|
{
|
|
RecursiveGuard l(x_sessions);
|
|
for (auto p: m_peers)
|
|
if (p.second->shouldReconnect())
|
|
toConnect.push_back(p.second);
|
|
}
|
|
|
|
for (auto p: toConnect)
|
|
if (openSlots--)
|
|
connect(p);
|
|
else
|
|
break;
|
|
|
|
m_nodeTable->discover();
|
|
}
|
|
|
|
auto runcb = [this](boost::system::error_code const& error) { run(error); };
|
|
m_timer->expires_from_now(boost::posix_time::milliseconds(c_timerInterval));
|
|
m_timer->async_wait(runcb);
|
|
}
|
|
|
|
void Host::startedWorking()
|
|
{
|
|
asserts(!m_timer);
|
|
|
|
{
|
|
// prevent m_run from being set to true at same time as set to false by stop()
|
|
// don't release mutex until m_timer is set so in case stop() is called at same
|
|
// time, stop will wait on m_timer and graceful network shutdown.
|
|
Guard l(x_runTimer);
|
|
// create deadline timer
|
|
m_timer.reset(new boost::asio::deadline_timer(m_ioService));
|
|
m_run = true;
|
|
}
|
|
|
|
// try to open acceptor (todo: ipv6)
|
|
m_listenPort = Network::tcp4Listen(m_tcp4Acceptor, m_netPrefs.listenPort);
|
|
|
|
// start capability threads
|
|
for (auto const& h: m_capabilities)
|
|
h.second->onStarting();
|
|
|
|
// determine public IP, but only if we're able to listen for connections
|
|
// todo: GUI when listen is unavailable in UI
|
|
if (m_listenPort)
|
|
{
|
|
determinePublic(m_netPrefs.publicIP, m_netPrefs.upnp);
|
|
|
|
if (m_listenPort > 0)
|
|
runAcceptor();
|
|
}
|
|
else
|
|
clog(NetNote) << "p2p.start.notice id:" << id().abridged() << "Listen port is invalid or unavailable. Node Table using default port (30303).";
|
|
|
|
// TODO: add m_tcpPublic endpoint; sort out endpoint stuff for nodetable
|
|
m_nodeTable.reset(new NodeTable(m_ioService, m_alias, m_listenPort > 0 ? m_listenPort : 30303));
|
|
m_nodeTable->setEventHandler(new HostNodeTableHandler(*this));
|
|
restoreNetwork(&m_restoreNetwork);
|
|
|
|
clog(NetNote) << "p2p.started id:" << id().abridged();
|
|
|
|
run(boost::system::error_code());
|
|
}
|
|
|
|
void Host::doWork()
|
|
{
|
|
if (m_run)
|
|
m_ioService.run();
|
|
}
|
|
|
|
void Host::keepAlivePeers()
|
|
{
|
|
if (chrono::steady_clock::now() - c_keepAliveInterval < m_lastPing)
|
|
return;
|
|
|
|
RecursiveGuard l(x_sessions);
|
|
for (auto p: m_sessions)
|
|
if (auto pp = p.second.lock())
|
|
pp->ping();
|
|
|
|
m_lastPing = chrono::steady_clock::now();
|
|
}
|
|
|
|
void Host::disconnectLatePeers()
|
|
{
|
|
auto now = chrono::steady_clock::now();
|
|
if (now - c_keepAliveTimeOut < m_lastPing)
|
|
return;
|
|
|
|
RecursiveGuard l(x_sessions);
|
|
for (auto p: m_sessions)
|
|
if (auto pp = p.second.lock())
|
|
if (now - c_keepAliveTimeOut > m_lastPing && pp->m_lastReceived < m_lastPing)
|
|
pp->disconnect(PingTimeout);
|
|
}
|
|
|
|
bytes Host::saveNetwork() const
|
|
{
|
|
if (!m_nodeTable)
|
|
return bytes();
|
|
|
|
std::list<Peer> peers;
|
|
{
|
|
RecursiveGuard l(x_sessions);
|
|
for (auto p: m_peers)
|
|
if (p.second)
|
|
peers.push_back(*p.second);
|
|
}
|
|
peers.sort();
|
|
|
|
RLPStream network;
|
|
int count = 0;
|
|
{
|
|
RecursiveGuard l(x_sessions);
|
|
for (auto const& p: peers)
|
|
{
|
|
// TODO: alpha: Figure out why it ever shares these ports.//p.address.port() >= 30300 && p.address.port() <= 30305 &&
|
|
// TODO: alpha: if/how to save private addresses
|
|
// Only save peers which have connected within 2 days, with properly-advertised port and public IP address
|
|
if (chrono::system_clock::now() - p.m_lastConnected < chrono::seconds(3600 * 48) && p.peerEndpoint().port() > 0 && p.peerEndpoint().port() < /*49152*/32768 && p.id != id() && !isPrivateAddress(p.peerEndpoint().address()))
|
|
{
|
|
network.appendList(10);
|
|
if (p.peerEndpoint().address().is_v4())
|
|
network << p.peerEndpoint().address().to_v4().to_bytes();
|
|
else
|
|
network << p.peerEndpoint().address().to_v6().to_bytes();
|
|
// TODO: alpha: replace 0 with trust-state of node
|
|
network << p.peerEndpoint().port() << p.id << 0
|
|
<< chrono::duration_cast<chrono::seconds>(p.m_lastConnected.time_since_epoch()).count()
|
|
<< chrono::duration_cast<chrono::seconds>(p.m_lastAttempted.time_since_epoch()).count()
|
|
<< p.m_failedAttempts << (unsigned)p.m_lastDisconnect << p.m_score << p.m_rating;
|
|
count++;
|
|
}
|
|
}
|
|
}
|
|
|
|
if (!!m_nodeTable)
|
|
{
|
|
auto state = m_nodeTable->snapshot();
|
|
state.sort();
|
|
for (auto const& s: state)
|
|
{
|
|
network.appendList(3);
|
|
if (s.endpoint.tcp.address().is_v4())
|
|
network << s.endpoint.tcp.address().to_v4().to_bytes();
|
|
else
|
|
network << s.endpoint.tcp.address().to_v6().to_bytes();
|
|
network << s.endpoint.tcp.port() << s.id;
|
|
count++;
|
|
}
|
|
}
|
|
|
|
RLPStream ret(3);
|
|
ret << 1 << m_alias.secret();
|
|
ret.appendList(count).appendRaw(network.out(), count);
|
|
return ret.out();
|
|
}
|
|
|
|
void Host::restoreNetwork(bytesConstRef _b)
|
|
{
|
|
// nodes can only be added if network is added
|
|
if (!isStarted())
|
|
BOOST_THROW_EXCEPTION(NetworkStartRequired());
|
|
|
|
RecursiveGuard l(x_sessions);
|
|
RLP r(_b);
|
|
if (r.itemCount() > 0 && r[0].isInt() && r[0].toInt<int>() == 1)
|
|
{
|
|
// r[0] = version
|
|
// r[1] = key
|
|
// r[2] = nodes
|
|
|
|
for (auto i: r[2])
|
|
{
|
|
bi::tcp::endpoint tcp;
|
|
bi::udp::endpoint udp;
|
|
if (i[0].itemCount() == 4)
|
|
{
|
|
tcp = bi::tcp::endpoint(bi::address_v4(i[0].toArray<byte, 4>()), i[1].toInt<short>());
|
|
udp = bi::udp::endpoint(bi::address_v4(i[0].toArray<byte, 4>()), i[1].toInt<short>());
|
|
}
|
|
else
|
|
{
|
|
tcp = bi::tcp::endpoint(bi::address_v6(i[0].toArray<byte, 16>()), i[1].toInt<short>());
|
|
udp = bi::udp::endpoint(bi::address_v6(i[0].toArray<byte, 16>()), i[1].toInt<short>());
|
|
}
|
|
auto id = (NodeId)i[2];
|
|
if (i.itemCount() == 3)
|
|
m_nodeTable->addNode(id, udp, tcp);
|
|
else if (i.itemCount() == 10)
|
|
{
|
|
shared_ptr<Peer> p = make_shared<Peer>();
|
|
p->id = id;
|
|
p->m_lastConnected = chrono::system_clock::time_point(chrono::seconds(i[4].toInt<unsigned>()));
|
|
p->m_lastAttempted = chrono::system_clock::time_point(chrono::seconds(i[5].toInt<unsigned>()));
|
|
p->m_failedAttempts = i[6].toInt<unsigned>();
|
|
p->m_lastDisconnect = (DisconnectReason)i[7].toInt<unsigned>();
|
|
p->m_score = (int)i[8].toInt<unsigned>();
|
|
p->m_rating = (int)i[9].toInt<unsigned>();
|
|
p->endpoint.tcp = tcp;
|
|
p->endpoint.udp = udp;
|
|
m_peers[p->id] = p;
|
|
m_nodeTable->addNode(*p.get());
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
KeyPair Host::networkAlias(bytesConstRef _b)
|
|
{
|
|
RLP r(_b);
|
|
if (r.itemCount() == 3 && r[0].isInt() && r[0].toInt<int>() == 1)
|
|
return move(KeyPair(move(Secret(r[1].toBytes()))));
|
|
else
|
|
return move(KeyPair::create());
|
|
}
|
|
|