Fields

• dtTraceLocalRootPeersTracer :: Tracer m (TraceLocalRootPeers ntnAddr resolverError)
   
• dtTracePublicRootPeersTracer :: Tracer m TracePublicRootPeers
   
• dtTraceLedgerPeersTracer :: Tracer m TraceLedgerPeers

  Ledger Peers tracer

• dtTracePeerSelectionTracer :: Tracer m (TracePeerSelection ntnAddr)
   
• dtDebugPeerSelectionInitiatorTracer :: Tracer m (DebugPeerSelection ntnAddr)
   
• dtDebugPeerSelectionInitiatorResponderTracer :: Tracer m (DebugPeerSelection ntnAddr)
   
• dtTracePeerSelectionCounters :: Tracer m PeerSelectionCounters
   
• dtTraceChurnCounters :: Tracer m ChurnCounters
   
• dtPeerSelectionActionsTracer :: Tracer m (PeerSelectionActionsTrace ntnAddr ntnVersion)
   
• dtConnectionManagerTracer :: Tracer m (Trace ntnAddr (ConnectionHandlerTrace ntnVersion ntnVersionData))
   
• dtConnectionManagerTransitionTracer :: Tracer m (AbstractTransitionTrace ConnStateId)
   
• dtServerTracer :: Tracer m (Trace ntnAddr)
   
• dtInboundGovernorTracer :: Tracer m (Trace ntnAddr)
   
• dtInboundGovernorTransitionTracer :: Tracer m (RemoteTransitionTrace ntnAddr)
   
• dtLocalConnectionManagerTracer :: Tracer m (Trace ntcAddr (ConnectionHandlerTrace ntcVersion ntcVersionData))

  Connection manager tracer for local clients

• dtLocalServerTracer :: Tracer m (Trace ntcAddr)

  Server tracer for local clients

• dtLocalInboundGovernorTracer :: Tracer m (Trace ntcAddr)

  Inbound protocol governor tracer for local clients

Fields

• daPeerTargets :: ConsensusModePeerTargets

  selection targets for the peer governor

• daReadLocalRootPeers :: STM m (Config RelayAccessPoint)
   
• daReadPublicRootPeers :: STM m (Map RelayAccessPoint PeerAdvertise)
   
• daConsensusMode :: ConsensusMode

  When syncing up, ie. ledgerStateJudgement == TooOld, when this is True we will maintain connection with many big ledger peers to get a strong guarantee that when syncing up we will finish with a true ledger state. When false, we will fall back on the previous algorithms that leverage UseBootstrapPeers flag

• daMinBigLedgerPeersForTrustedState :: MinBigLedgerPeersForTrustedState

  For Genesis, this sets the floor for minimum number of active big ledger peers we must be connected to in order to be able to signal trusted state (OutboundConnectionsState)

• daReadUseBootstrapPeers :: STM m UseBootstrapPeers
   
• daReadLedgerPeerSnapshot :: STM m (Maybe LedgerPeerSnapshot)

  Depending on configuration, node may provide us with a snapshot of big ledger peers taken at some slot on the chain. These peers may be selected by ledgerPeersThread when requested by the peer selection governor when the node is syncing up. This is especially useful for Genesis consensus mode.

• daOwnPeerSharing :: PeerSharing

  Peer's own PeerSharing value.

  This value comes from the node's configuration file and is static.

• daReadUseLedgerPeers :: STM m UseLedgerPeers
   
• daProtocolIdleTimeout :: DiffTime

  Timeout which starts once all responder protocols are idle. If the responders stay idle for duration of the timeout, the connection will be demoted, if it wasn't used by the p2p-governor it will be closed.

  Applies to Unidirectional as well as Duplex node-to-node connections.

  See serverProtocolIdleTimeout.

• daTimeWaitTimeout :: DiffTime

  Time for which node-to-node connections are kept in TerminatingState, it should correspond to the OS configured TCP TIME_WAIT timeout.

  This timeout will apply to after a connection has been closed, its purpose is to be resilient for delayed packets in the same way TCP is using TIME_WAIT.

• daDeadlineChurnInterval :: DiffTime

  Churn interval between churn events in deadline mode. A small fuzz is added (max 10 minutes) so that not all nodes churn at the same time.

  By default it is set to 3300 seconds.

• daBulkChurnInterval :: DiffTime

  Churn interval between churn events in bulk sync mode. A small fuzz is added (max 1 minute) so that not all nodes churn at the same time.

  By default it is set to 300 seconds.

Fields

• acceptedConnectionsHardLimit :: !Word32

  Hard limit of accepted connections.

• acceptedConnectionsSoftLimit :: !Word32

  Soft limit of accepted connections. If we are above this threshold, we will start rate limiting.

• acceptedConnectionsDelay :: !DiffTime

  Max delay for limiting accepted connections. We use linear regression starting from 0 at the soft limit up to acceptedConnectionDelay at the hard limit.

Fields

• daRethrowPolicy :: RethrowPolicy

  node-to-node rethrow policy

• daReturnPolicy :: ReturnPolicy a

  node-to-node return policy

• daLocalRethrowPolicy :: RethrowPolicy

  node-to-client rethrow policy

• daPeerMetrics :: PeerMetrics m ntnAddr

  PeerMetrics used by peer selection policy (see simplePeerSelectionPolicy)

• daBlockFetchMode :: STM m FetchMode

  Used by churn-governor

• daPeerSharingRegistry :: PeerSharingRegistry ntnAddr m

  Used for peer sharing protocol

Fields

• diNtnSnocket :: Snocket m ntnFd ntnAddr

  node-to-node snocket

• diNtnBearer :: MakeBearer m ntnFd

  node-to-node MakeBearer callback

• diNtnConfigureSocket :: ntnFd -> Maybe ntnAddr -> m ()

  node-to-node socket configuration

  It is used by both inbound and outbound connection. The address is the local address that we can bind to if given (NOTE: for node-to-node connection Just is always given).

• diNtnConfigureSystemdSocket :: ntnFd -> ntnAddr -> m ()

  node-to-node systemd socket configuration

• diNtnHandshakeArguments :: HandshakeArguments (ConnectionId ntnAddr) ntnVersion ntnVersionData m

  node-to-node handshake configuration

• diNtnAddressType :: ntnAddr -> Maybe AddressType

  node-to-node address type

• diNtnDataFlow :: ntnVersionData -> DataFlow

  node-to-node data flow used by connection manager to classify negotiated connections

• diNtnPeerSharing :: ntnVersionData -> PeerSharing

  remote side peer sharing information used by peer selection governor to decide which peers are available for performing peer sharing

• diNtnToPeerAddr :: IP -> PortNumber -> ntnAddr

  node-to-node peer address

• diNtcSnocket :: Snocket m ntcFd ntcAddr

  node-to-client snocket

• diNtcBearer :: MakeBearer m ntcFd

  node-to-client MakeBearer callback

• diNtcHandshakeArguments :: HandshakeArguments (ConnectionId ntcAddr) ntcVersion ntcVersionData m

  node-to-client handshake configuration

• diNtcGetFileDescriptor :: ntcFd -> m FileDescriptor

  node-to-client file descriptor

• diRng :: StdGen

  diffusion pseudo random generator. It is split between various components that need randomness, e.g. inbound governor, peer selection, policies, etc.

• diInstallSigUSR1Handler :: forall (mode :: Mode) x y. NodeToNodeConnectionManager mode ntnFd ntnAddr ntnVersionData ntnVersion m x y -> StrictTVar m (PeerSelectionState ntnAddr (NodeToNodePeerConnectionHandle mode ntnAddr ntnVersionData m x y)) -> PeerMetrics m ntnAddr -> m ()

  callback which is used to register SIGUSR1 signal handler.

• diDnsActions :: DNSLookupType -> DNSActions resolver resolverError m

  diffusion dns actions

• diUpdateVersionData :: ntnVersionData -> DiffusionMode -> ntnVersionData

  Update ntnVersionData for initiator-only local roots.

• diConnStateIdSupply :: ConnStateIdSupply m

  ConnStateIdSupply used by the connection-manager for this node.

  This is exposed for testing, where we use a global ConnStateIdSupply.