Tests if addresses are online or offline by briefly connecting to them. These short lived connections are referred to as feeler connections. Feeler connections are designed to increase the number of fresh online addresses in tried by selecting and connecting to addresses in new. One feeler connection is attempted on average once every two minutes.
This change was suggested as Countermeasure 4 in
Eclipse Attacks on Bitcoin’s Peer-to-Peer Network, Ethan Heilman,
Alison Kendler, Aviv Zohar, Sharon Goldberg. ePrint Archive Report
2015/263. March 2015.
* Rework addnode behaviour
* Use CNode::addeName to track whether a connection to a name is already open
* A new connection to a previously-connected by-name addednode is only opened when
the previous one closes (even if the name starts resolving to something else)
* At most one connection is opened per addednode (even if the name resolves to multiple)
* Unify the code between ThreadOpenAddedNodeConnections and getaddednodeinfo
* Information about open connections is always returned, and the dns argument becomes a dummy
* An IP address and inbound/outbound is only reported for the (at most 1) open connection
* Prevent duplicate connections where one is by name and another by ip
* Randomize name lookup result in ConnectSocketByName
This reverts commit 1f828f45ec.
The commit being reverted changed FindNode(const CService& addr)
to make no difference between nodes with the same IP address,
but different ports, but only for regtest network.
As functional tests run several nodes on different ports or the same
IP address (127.0.0.1), this eventually started breaking functional tests.
The only use for regtest network I know is for functional tests,
so it's time to revert that commit.
* Add recently accepted blocks and txn to AttemptToEvictConnection.
This protects any not-already-protected peers who were the most
recent four to relay transactions and most recent four to send
blocks to us.
* Allow disconnecting a netgroup with only one member in eviction.
With the latest additions there are enough protective measures that
we can take the training wheels off.
* net: don't import std namespace
This file is about to be broken up into chunks and moved around. Drop the
namespace now rather than requiring other files to use it.
* net: remove unused set
* net: use the exposed GetNodeSignals() rather than g_signals directly
* net: make Ban/Unban/ClearBan functionality consistent
- Ban/Unban/ClearBan call uiInterface.BannedListChanged() as necessary
- Ban/Unban/ClearBan sync to disk if the operation is user-invoked
- Mark node for disconnection automatically when banning
- Lock cs_vNodes while setting disconnected
- Don't spin in a tight loop while setting disconnected
* net: No need to export DumpBanlist
DumpBanList currently does this:
- with lock: take a copy of the banmap
- perform I/O (write out the banmap)
- with lock: mark the banmap non-dirty
If a new ban is added during the I/O operation, it may never be persisted to
disk.
Reorder operations so that the data to be persisted cannot be older than the
time at which the banmap was marked non-dirty.
* Only store and connect to NODE_NETWORK nodes
* Keep addrman's nService bits consistent with outbound observations
* Verify that outbound connections have expected services
* Don't require services in -addnode
* Introduce enum ServiceFlags for service flags
* Introduce REQUIRED_SERVICES constant
* CAddrDB modified so that when de-serialization code throws an exception Addrman is reset to a clean state
* CAddrDB modified to make unit tests possible
* Regression test created to ensure bug is fixed
* StartNode modifed to clear adrman if CAddrDB::Read returns an error code.
* banlist: update set dirty to be more fine grained
- move the SetBannedSetDirty(false) call from DumpData() into DumpBanlist()
- ensure we only set false, if the write succeeded
* banlist: better handling of banlist in StartNode()
- only start working on/with banlist data, if reading in the banlist from
disk didn't fail
- as CNode::setBannedIsDirty is false (default) when reading fails, we
don't need to explicitly set it to false to prevent writing
banlist.dat in that case either
* banlist: add more banlist infos to log / add GUI signal
- to match the peers.dat handling also supply a debug.log entry for how
many entries were loaded from banlist.dat and how long it took
- add a GUI init message for loading the banlist (same as with peers.dat)
- move the same message for peers.dat upwards in the code, to be able to
reuse the timing variable nStart and also just log, if our read from
peers.dat didn't fail
* banlist (bugfix): allow CNode::SweepBanned() to run on interval
- allows CNode::SweepBanned() to run, even if !CNode::BannedSetIsDirty(),
because if nBanUntil is over we want the ban to be disabled for these
nodes
* log bytes recv/sent per command
* net: Account for `sendheaders` `verack` messages
Looks like these were forgotten in #6589.
* Backport remaining part of Bitcoin PR bitcoin/bitcoin#7181.
Most of this PR is already merged, but a small part remaining
that makes per-command byte counts in CNode working.
Signed-off-by: Oleg Girko <ol@infoserver.lv>
* Few networking fixes:
- skip "masternode"/inbound connections for sync related processes
- do not sync gov data to other nodes until fully synced ourselves
- do not accept incoming connections until fully synced
* inbound connections could be harmful only if our node is a masternode
* same for CGovernanceManager::Sync
* Multi-quorum InstantSend, complete refactoring
+ cleanup for IS and partial protobump
* more changes:
- allow InstantSend tx to have 10 inputs max
- store many unique tx hashes in mapVotedOutpoints
- more checks in AcceptToMemoryPoolWorker (moved from ProcessMessage + CTxLockRequest(tx).IsValid() )
* More changes:
- let multiple lock candidates compete for votes
- fail to vote on the same outpoint twice early
* More changes:
- notify CInstantSend on UpdatedBlockTip -> remove cs_main from CheckAndRemove()
- notify CInstantSend on SyncTransaction -> count expiration block starting from the block corresponding tx was confirmed instead of the block lock candidate/vote was created
- fixed few locks
* add comments about nConfirmedHeight
* Fix "Block vs Lock" edge case
* Fix "Block vs Lock" edge case, p2
* Fix issues:
- fix logic for locking inputs and notifying - see UpdateLockedTransaction, TryToFinalizeLockCandidate
- add missing hash inserting in ProcessTxLockVote
- add nMaxBlocks param to ResolveConflicts to limit max depth allowed to disconnect blocks recursively
- fix false positive mempool conflict
- add missing mutex locks
- fix fRequireUnspent logic in CTxLockRequest::IsValid
- some were not used, some were included twice, some were in the wrong place, some were missing (but it compiled because some were in the wrong place)
- organized a bit better, grouped dash specific includes in original bitcoin files, should save some time solving conflicts when/if merging patches later
86d8505 Refactor CActiveMasternode
+ move strMasterNodeAddr to CActiveMasternode
a005c79 Refactor InstantSend
+ new lock cs_instantsend to protect maps on CleanTransactionLocksList()
+ new DEFAULT_INSTANTSEND_DEPTH constant
+ rename MIN_INSTANTX_PROTO_VERSION to MIN_INSTANTSEND_PROTO_VERSION and bump it
d24182c Refactor Privatesend
+ decouple from util.h and version.h
+ more functions for CDarksendBroadcastTx: constructors, signing, serialization
+ move from rand() to insecure_rand() in general but to GetRand() for session id
+ fix defaults
513506f Fixing AddRef() usage
Using AddRef() in ConnectNode() for existing connections doesn't feel right considering how refs are released in ThreadSocketHandler(). I guess this could be the reason that sometimes refs stay >0 no matter what and nodes stuck in vNodesDisconnected forever which means that node never get deleted and FinalizeNode signal is never fired which in its turn means that for example mapBlocksInFlight can't be cleaned properly and then blocks stuck.
This commit should solve the issue by:
- removing AddRef() for existing connections
- adding AddRef() in CNode's constructor using the same conditions as in ThreadSocketHandler()
- addding AddRef() in ConnectNode() and Release() in ThreadSocketHandler() for mixing nodes
- removing explicit calls to Release() (back to `pnode->fDisconnect = true` in `CMasternodeMan::ProcessMasternodeConnections`)
9da4a83 fix names/comments
487674f Governance object/vote syncing fixes
- disable fCached values
- use two maps for storing votes, by hash and parent-hash/type
- disable part of flatdb.dump (still overwriting)
- fixed govobj/votes relay and sync
15821fe various fixes
- Added const where possible
- Uncommented sync block
- Protocol min 70201
- Fixed bug which flags invalid votes incorrectly
- Formatting
aa8fdd7 fix curly braces
d8e39b1 Fix GetTypeHash bug
- Should not collide based on the outcome
732a8a3 fixed mismatched index for vote map
Previously we used the CInv that would be sent to the peer announcing the
transaction as the key, but using the txid instead allows us to decouple the
p2p layer from the application logic (which relies on this map to avoid
duplicate tx requests).
Github-Pull: #7862
Rebased-From: 7e91f632c7
- Fix CreateTransaction and GetDepthInMainChain for IX compatibility
- Prepare IsIXTXValid for the next update (see FIXME in CreateTransaction)
- Unify processing of TX/IX/DSTX a bit more
- Clean up some code
This corrects a bug the case of tying group size where the code may
fail to select the group with the newest member. Since newest time
is the final selection criteria, failing to break ties on it
on the step before can undermine the final selection.
Tied netgroups are very common.
With automatic tor HS support in place we should probably not be providing
absolute protection for local peers, since HS inbound could be used to
attack pretty easily. Instead, this counts on the latency metric inside
AttemptToEvictConnection to privilege actually local peers.
When running the rpc tests in Wine, nodes often fail to listen on localhost
due to a stale socket from a previous run. This aligns the behavior with other
platforms.
We used to have a trickle node, a node which was chosen in each iteration of
the send loop that was privileged and allowed to send out queued up non-time
critical messages. Since the removal of the fixed sleeps in the network code,
this resulted in fast and attackable treatment of such broadcasts.
This pull request changes the 3 remaining trickle use cases by random delays:
* Local address broadcast (while also removing the the wiping of the seen filter)
* Address relay
* Inv relay (for transactions; blocks are always relayed immediately)
The code is based on older commits by Patrick Strateman.
Github-Pull: #7125
Rebased-From: 5400ef6bcb
- Avoids string typos (by making the compiler check)
- Makes it easier to grep for handling/generation of a certain message type
- Refer directly to documentation by following the symbol in IDE
- Move list of valid message types to protocol.cpp:
protocol.cpp is a more appropriate place for this, and having
the array there makes it easier to keep things consistent.
Github-Pull: #7181
Rebased-From: 9bbe71b641
ebb25f4 Limit setAskFor and retire requested entries only when a getdata returns. (Gregory Maxwell)
5029698 prevent peer flooding request queue for an inv (kazcw)
The setAskFor duplicate elimination was too eager and removed entries
when we still had no getdata response, allowing the peer to keep
INVing and not responding.
mapAlreadyAskedFor does not keep track of which peer has a request queued for a
particular tx. As a result, a peer can blind a node to a tx indefinitely by
sending many invs for the same tx, and then never replying to getdatas for it.
Each inv received will be placed 2 minutes farther back in mapAlreadyAskedFor,
so a short message containing 10 invs would render that tx unavailable for 20
minutes.
This is fixed by disallowing a peer from having more than one entry for a
particular inv in mapAlreadyAskedFor at a time.
- Force AUTHCOOKIE size to be 32 bytes: This provides protection against
an attack where a process pretends to be Tor and uses the cookie
authentication method to nab arbitrary files such as the
wallet
- torcontrol logging
- fix cookie auth
- add HASHEDPASSWORD auth, fix fd leak when fwrite() fails
- better error reporting when cookie file is not ok
- better init/shutdown flow
- stop advertizing service when disconnected from tor control port
- COOKIE->SAFECOOKIE auth
Starting with Tor version 0.2.7.1 it is possible, through Tor's control socket
API, to create and destroy 'ephemeral' hidden services programmatically.
https://stem.torproject.org/api/control.html#stem.control.Controller.create_ephemeral_hidden_service
This means that if Tor is running (and proper authorization is available),
bitcoin automatically creates a hidden service to listen on, without user
manual configuration. This will positively affect the number of available
.onion nodes.
- When the node is started, connect to Tor through control socket
- Send `ADD_ONION` command
- First time:
- Make it create a hidden service key
- Save the key in the data directory for later usage
- Make it redirect port 8333 to the local port 8333 (or whatever port we're listening on).
- Keep control socket connection open for as long node is running. The hidden service will
(by default) automatically go away when the connection is closed.
* -maxuploadtarget can be set in MiB
* if <limit> - ( time-left-in-24h-cycle / 600 * MAX_BLOCK_SIZE ) has reach, stop serve blocks older than one week and filtered blocks
* no action if limit has reached, no guarantee that the target will not be surpassed
* add outbound limit informations to rpc getnettotals
Nagle appears to be a significant contributor to latency now that the static
sleeps are gone. Most of our messages are relatively large compared to
IP + TCP so I do not expect this to create enormous overhead.
This may also reduce traffic burstyness somewhat.
`nMinPingUsecTime` was left uninitialized in CNode.
The correct initialization for a minimum-until-now is int64_t's max value, so initialize it to that.
Thanks @MarcoFalke for noticing.
027de94 Use network group instead of CNetAddr in final pass to select node to disconnect (Patrick Strateman)
000c18a Fix comment (Patrick Strateman)
fed3094 Acquire cs_vNodes before changing refrence counts (Patrick Strateman)
69ee1aa CNodeRef copy constructor and assignment operator (Patrick Strateman)
dc81dd0 Return false early if vEvictionCandidates is empty (Patrick Strateman)
17f3533 Better support for nodes with non-standard nMaxConnections (Patrick Strateman)
1317cd1 RAII wrapper for CNode* (Patrick Strateman)
df23937 Add comments to AttemptToEvictConnection (Patrick Strateman)
a8f6e45 Remove redundant whiteconnections option (Patrick Strateman)
b105ba3 Prefer to disconnect peers in favor of whitelisted peers (Patrick Strateman)
2c70153 AttemptToEvictConnection (Patrick Strateman)
4bac601 Record nMinPingUsecTime (Patrick Strateman)
ae037b7 Refactor: Move failure conditions to the top of AcceptConnection (Patrick Strateman)
1ef4817 Refactor: Bail early in AcceptConnection (Patrick Strateman)
541a1dd Refactor: AcceptConnection (Patrick Strateman)
When running the rpc tests in Wine, nodes often fail to listen on localhost
due to a stale socket from a previous run. This aligns the behavior with other
platforms.
7b79cbd limit total length of user agent comments (Pavol Rusnak)
557f8ea implement uacomment config parameter which can add comments to user agent as per BIP-0014 (Pavol Rusnak)
While CBloomFilter is usually used with an explicitly set nTweak,
CRollingBloomFilter is only used internally. Requiring every caller to
set nTweak is error-prone and redundant; better to have the class handle
that for you with a high-quality randomness source.
Additionally when clearing the filter it makes sense to change nTweak as
well to recover from a bad setting, e.g. due to insufficient randomness
at initialization, so the clear() method is replaced by a reset() method
that sets a new, random, nTweak value.
a794284 locking: add a quick example of GUARDED_BY (Cory Fields)
2b890dd locking: fix a few small issues uncovered by -Wthread-safety (Cory Fields)
cd27bba locking: teach Clang's -Wthread-safety to cope with our scoped lock macros (Cory Fields)
Submissions to the network now require a fee to be paid to the network (mining fee) using a special transaction with a OP_RETURN && ProposalHash in one of the outputs. This allows the network to filter spam quickly, while also allowing anyone to submit a proposal to the network.
To implement these changes we've introduced a few new commands:
mnbudget prepare PROPOSAL-NAME URL PAYMENT_COUNT BLOCK_START DASH_ADDRESS DASH_AMOUNT YES|NO|ABSTAIN [USE_IX(TRUE|FALSE)]
- To create the special transaction
mnbudget submit PROPOSAL-NAME URL PAYMENT_COUNT BLOCK_START DASH_ADDRESS DASH_AMOUNT YES|NO|ABSTAIN FEE_TX
- After the transaction is accepted by the network and has 3 confirmations, you can submit the transaction to the network here
mnbudget show
- Get the proposal hash from here
mnbudget vote PROPOSAL-HASH YES|NO|ABSTAIN
- You can now simply vote by hash using this command
- rpcwallet: No need to lock twice here
- openssl: Clang doesn't understand selective lock/unlock here. Ignore it.
- CNode: Fix a legitimate (though very unlikely) locking bug.
This sets aside a number of connection slots for whitelisted peers,
useful for ensuring your local users and miners can always get in,
even if your limit on inbound connections has already been reached.
- Added commands for using budgets "mnbudget" and "mnfinalbudget"
- Supports 100% decentralized budget control and view-only site with json meta data object
Use a probabilistic bloom filter to keep track of which addresses
we think we have given our peers, instead of a list.
This uses much less memory, at the cost of sometimes failing to
relay an address to a peer-- worst case if the bloom filter happens
to be as full as it gets, 1-in-1,000.
Measured memory usage of a full mruset setAddrKnown: 650Kbytes
Constant memory usage of CRollingBloomFilter addrKnown: 37Kbytes.
This will also help heap fragmentation, because the 37K of storage
is allocated when a CNode is created (when a connection to a peer
is established) and then there is no per-item-remembered memory
allocation.
I plan on testing by restarting a full node with an empty peers.dat,
running a while with -debug=addrman and -debug=net, and making sure
that the 'addr' message traffic out is reasonable.
(suggestions for better tests welcome)
1d21ba2 Scale up addrman (Pieter Wuille)
c6a63ce Always use a 50% chance to choose between tried and new entries (Pieter Wuille)
f68ba3f Do not bias outgoing connections towards fresh addresses (Pieter Wuille)
a8ff7c6 Simplify hashing code (Pieter Wuille)
e6b343d Make addrman's bucket placement deterministic. (Pieter Wuille)
b23add5 Switch addrman key from vector to uint256 (Pieter Wuille)
This change was suggested as Countermeasure 2 in
Eclipse Attacks on Bitcoin’s Peer-to-Peer Network, Ethan Heilman,
Alison Kendler, Aviv Zohar, Sharon Goldberg. ePrint Archive Report
2015/263. March 2015.
Normally bitcoin core does not display any network originated strings without
sanitizing or hex encoding. This wasn't done for strcommand in many places.
This could be used to play havoc with a terminal displaying the logs,
especially with printtoconsole in use.
Thanks to Evil-Knievel for reporting this issue.
Conflicts:
src/main.cpp
Normally bitcoin core does not display any network originated strings without
sanitizing or hex encoding. This wasn't done for strcommand in many places.
This could be used to play havoc with a terminal displaying the logs,
especially with printtoconsole in use.
Thanks to Evil-Knievel for reporting this issue.
856e862 namespace: drop most boost namespaces and a few header cleanups (Cory Fields)
9b1ab86 namespace: drop boost::assign altogether here (Cory Fields)
a324199 namespace: remove boost namespace pollution (Cory Fields)
- Made masternodes/darksend compatible with regression testing mode (a local-only blockchain that doesn't require mining). Developers can now test multiple rounds in a few minutes without waiting on mining (much faster).
- Added dsee security verification to v11
- darkSendMasternodes -> vecMasternodes (must clearer)
This avoids connecting to them again too soon in ThreadOpenConnections.
Make an exception for connection failures to the proxy as these
shouldn't affect the status of specific nodes.
This is a simplified re-do of closed pull #3088.
This patch eliminates the privacy and reliability problematic use
of centralized web services for discovering the node's addresses
for advertisement.
The Bitcoin protocol already allows your peers to tell you what
IP they think you have, but this data isn't trustworthy since
they could lie. So the challenge is using it without creating a
DOS vector.
To accomplish this we adopt an approach similar to the one used
by P2Pool: If we're announcing and don't have a better address
discovered (e.g. via UPNP) or configured we just announce to
each peer the address that peer told us. Since peers could
already replace, forge, or drop our address messages this cannot
create a new vulnerability... but if even one of our peers is
giving us a good address we'll eventually make a useful
advertisement.
We also may randomly use the peer-provided address for the
daily rebroadcast even if we otherwise have a seemingly routable
address, just in case we've been misconfigured (e.g. by UPNP).
To avoid privacy problems, we only do these things if discovery
is enabled.
Many changes:
* Do not use 'getblocks', but 'getheaders', and use it to build a headers tree.
* Blocks are fetched in parallel from all available outbound peers, using a
limited moving window. When one peer stalls the movement of the window, it is
disconnected.
* No more orphan blocks. At all. We only ever request a block for which we have
verified the headers, and store it to disk immediately. This means that a
disk-fill attack would require PoW.
* Require protocol version 31800 for every peer (released in december 2010).
* No more syncnode (we sync from everyone we can, though limited to 1 during
initial *headers* sync).
* Introduce some extra named constants, comments and asserts.
Avoids that SOCKS5 negotiation will hold up the shutdown process.
- Sockets can stay in non-blocking mode, no need to switch it on/off
anymore
- Adds a timeout (20 seconds) on SOCK5 negotiation. This should be
enough for even Tor to get a connection to a hidden service, and
avoids blocking the opencon thread indefinitely on a hanging proxy.
Fixes#2954.
Split up util.cpp/h into:
- string utilities (hex, base32, base64): no internal dependencies, no dependency on boost (apart from foreach)
- money utilities (parsesmoney, formatmoney)
- time utilities (gettime*, sleep, format date):
- and the rest (logging, argument parsing, config file parsing)
The latter is basically the environment and OS handling,
and is stripped of all utility functions, so we may want to
rename it to something else than util.cpp/h for clarity (Matt suggested
osinterface).
Breaks dependency of sha256.cpp on all the things pulled in by util.
