hotstuff

README

Flow's HotStuff

We use a BFT consensus algorithm with deterministic finality in Flow for

• Consensus Nodes: deciding on the content of blocks
• Cluster of Collector Nodes: batching transactions into collections.

Flow uses a derivative of HotStuff (see paper HotStuff: BFT Consensus in the Lens of Blockchain (version 6) for the original algorithm). Our implementation is a mixture of Chained HotStuff and Event-Driven HotStuff with a few additional, minor modifications:

• We employ the rules for locking on blocks and block finalization from Event-Driven HotStuff. Specifically, we lock the newest block which has an (direct or indirect) 2-chain built on top. We finalize a block when it has a 3-chain built on top where with the first two chain links being direct.
• The way we progress though views follows the Chained HotStuff algorithm. A replica only votes or proposes for its current view.
• Flow's HotStuff does not implement ViewChange messages (akin to the approach taken in Streamlet: Textbook Streamlined Blockchains)
• Flow uses a decentralized random beacon (based on Dfinity's proposal). The random beacon is run by Flow's consensus nodes and integrated into the consensus voting process.

Node weights

In Flow, there are multiple HotStuff instances running in parallel. Specifically, the consensus nodes form a HotStuff committee and each collector cluster is its own committee. In the following, we refer to an authorized set of nodes, who run a particular HotStuff instance as a (HotStuff) committee.

• Flow allows nodes to have different weights, reflecting how much they are trusted by the protocol. The weight of a node can change over time due to stake changes or discovering protocol violations. A super-majority of nodes is defined as a subset of the consensus committee, where the nodes have more than 2/3 of the entire committee's accumulated weight.
• The messages from zero-weighted nodes are ignored by all committee members.

Architecture

Determining block validity

In addition to HotStuff's requirements on block validity, the Flow protocol adds additional requirements. For example, it is illegal to repeatedly include the same payload entities (e.g. collections, challenges, etc) in the same fork. Generally, payload entities expire. However within the expiry horizon, all ancestors of a block need to be known to verify that payload entities are not repeated.

We exclude the entire logic for determining payload validity from the HotStuff core implementation. This functionality is encapsulated in the Chain Compliance Layer (CCL) which precedes HotStuff. The CCL forwards a block to the HotStuff core logic only if

• the block's payload is valid,
• the block is connected to the most recently finalized block, and
• all ancestors have previously been processed by HotStuff.

If ancestors of a block are missing, the CCL caches the respective block and (iteratively) requests missing ancestors.

Payload generation

Payloads are generated outside of the HotStuff core logic. HotStuff only incorporates the payload root hash into the block header.

Structure of votes

In Flow's HotStuff implementation, votes are used for two purposes:

1. Prove that a super-majority of committee nodes considers the respective block a valid extension of the chain. Therefore, nodes include a StakingSignature (BLS with curve BLS12-381) in their vote.
2. Construct a Source of Randomness as described in Dfinity's Random Beacon. Therefore, consensus nodes include a RandomBeaconSignature (also BLS with curve BLS12-381, used in a threshold signature scheme) in their vote.

When the primary collects the votes, it verifies the StakingSignature and the RandomBeaconSignature. If either signature is invalid, the entire vote is discarded. From all valid votes, the StakingSignatures and the RandomBeaconSignatures are aggregated separately. (note: Only the aggregation of the RandomBeaconSignature into a threshold signature is currently implemented. Aggregation of the BLS StakingSignatures will be added later.)

• Following version 6 of the HotStuff paper, replicas forward their votes for block b to the leader of the next view, i.e. the primary for view b.View + 1.
• A proposer will attach its own vote for its proposal in the block proposal message (instead of signing the block proposal for authenticity and separately sending a vote).

Primary section

For primary section, we use a randomized, weight-proportional selection.

Component Overview

HotStuff's core logic is broken down into multiple components. The figure below illustrates the dependencies of the core components and information flow between these components.

• HotStuffEngine and ChainComplianceLayer do not contain any core HotStuff logic. They are responsible for relaying HotStuff messages and validating block payloads respectively.
• EventLoop buffers all incoming events, so EventHandler can process one event at a time in a single thread.
• EventHandler orchestrates all HotStuff components and implements HotStuff's state machine. The event handler is designed to be executed single-threaded.
• Communicator relays outgoing HotStuff messages (votes and block proposals)
• Pacemaker is a basic PaceMaker to ensure liveness by keeping the majority of committee replicas in the same view
• Forks maintains an in-memory representation of all blocks b, whose view is larger or equal to the view of the latest finalized block (known to this specific replica). As blocks with missing ancestors are cached outside of HotStuff (by the Chain Compliance Layer), all blocks stored in Forks are guaranteed to be connected to the genesis block (or the trusted checkpoint from which the replica started). Internally, Forks consists of multiple layers:
  • LevelledForest: A blockchain forms a Tree. When removing all blocks with views strictly smaller than the last finalized block, the chain decomposes into multiple disconnected trees. In graph theory, such structure is a forest. To separate general graph-theoretical concepts from the concrete block-chain application, LevelledForest refers to blocks as graph vertices and to a block's view number as level. The LevelledForest is an in-memory data structure to store and maintain a levelled forest. It provides functions to add vertices, query vertices by their ID (block's hash), query vertices by level, query the children of a vertex, and prune vertices by level (remove them from memory).
  • Finalizer: the finalizer uses the LevelledForest to store blocks. The Finalizer tracks the locked and finalized blocks. Furthermore, it evaluates whether a block is safe to vote for.
  • ForkChoice: implements the fork choice rule. Currently, NewestForkChoice always uses the quorum certificate (QC) with the largest view to build a new block (i.e. the fork a super-majority voted for most recently).
  • Forks is the highest level. It bundles the Finalizer and ForkChoice into one component.
• Validator validates the HotStuff-relevant aspects of
  • QC: total weight of all signers is more than 2/3 of committee weight, validity of signatures, view number is strictly monotonously increasing
  • block proposal: from designated primary for the block's respective view, contains proposer's vote for its own block, QC in block is valid
  • vote: validity of signature, voter is has positive weight
• VoteAggregator caches votes on a per-block basis and builds QC if enough votes have been accumulated.
• Voter tracks the view of the latest vote and determines whether or not to vote for a block (by calling forks.IsSafeBlock)
• Committee maintains the list of all authorized network members and their respective weight on a per-block basis. Furthermore, the committee contains the primary selection algorithm.
• BlockProducer constructs the payload of a block, after the HotStuff core logic has decided which fork to extend

Implementation

We have translated the Chained HotStuff protocol into a state machine shown below. The state machine is implemented in EventHandler.

PaceMaker

The HotStuff state machine interacts with the PaceMaker, which triggers view changes. Conceptually, the PaceMaker interfaces with the EventHandler in two different modes:

• [asynchronous] On timeouts, the PaceMaker will emit a timeout event, which is processed as any other event (such as incoming blocks or votes) through the EventLoop.
• [synchronous] When progress is made following the happy-path business logic, the EventHandler will inform the PaceMaker about completing the respective processing step via a direct method call (see PaceMaker interface). If the PaceMaker changed the view in response, it returns a NewViewEvent which will be synchronously processed by the EventHandler.

Flow's PaceMaker is simple (compared to Libra's PaceMaker for instance). It solely relies on increasing the timeouts exponentially if no progress is made and exponentially decreasing timeouts on the happy path. The timeout values are limited by lower and upper-bounds to ensure that the PaceMaker can change from large to small timeouts in a reasonable number of views. The specific values for lower and upper timeout bounds are protocol-specified; we envision the bounds to be on the order of 1sec (lower bound) and multiple days (upper bound).

Progress, from the perspective of the PaceMaker is defined as entering view V for which the replica knows a QC with V = QC.view + 1. In other words, we transition into the next view due to reaching quorum in the last view.

A central, non-trivial functionality of the PaceMaker is to skip views. Specifically, given a QC with view qc.view, the Pacemaker will skip ahead to view qc.view + 1 if currentView ≤ qc.view.

Code structure

All relevant code implementing the core HotStuff business logic is contained in /consensus/hotstuff/ (folder containing this README). When starting to look into the code, we suggest starting with /consensus/hotstuff/event_loop.go and /consensus/hotstuff/event_handler.go.

Folder structure

All files in the /consensus/hotstuff/ folder, except for event_loop.go and follower_loop.go, are interfaces for HotStuff-related components. The concrete implementations for all HotStuff-relevant components are in corresponding sub-folders. For completeness, we list the component implemented in each subfolder below:

• /consensus/hotstuff/blockproducer builds a block proposal for a specified QC, interfaces with the logic for assembling a block payload, combines all relevant fields into a new block proposal.
• /consensus/hotstuff/committee maintains the list of all authorized network members and their respective weight on a per-block basis; contains the primary selection algorithm.
• /consensus/hotstuff/eventhandler orchestrates all HotStuff components and implements the HotStuff state machine. The event handler is designed to be executed single-threaded.
• /consensus/hotstuff/follower This component is only used by nodes that are not participating in the HotStuff committee. As Flow has dedicated node roles with specialized network functions, only a subset of nodes run the full HotStuff protocol. Nevertheless, all nodes need to be able to act on blocks being finalized. The approach we have taken for Flow is that block proposals are broadcast to all nodes (including non-committee nodes). Non-committee nodes locally determine block finality by applying HotStuff's finality rules. The HotStuff Follower contains the functionality to consume block proposals and trigger downstream processing of finalized blocks. The Follower does not actively participate in HotStuff.
• /consensus/hotstuff/forks maintains an in-memory representation of all blocks b, whose view is larger or equal to the view of the latest finalized block (known to this specific HotStuff replica). It tracks the last finalized block, the currently locked block, evaluates whether it is safe to vote for a given block and provides a Fork-Choice when the replica is primary.
• /consensus/hotstuff/helper contains broadly-used helper functions for testing
• /consensus/hotstuff/integration integration tests for verifying correct interaction of multiple HotStuff replicas
• /consensus/hotstuff/model contains the HotStuff data models, including block proposal, QC, etc. Many HotStuff data models are built on top of basic data models defined in /model/flow/.
• /consensus/hotstuff/notifications: All relevant events within the HotStuff logic are exported though a notification system. While the notifications are not used HotStuff-internally, they notify other components within the same node of relevant progress and are used for collecting HotStuff metrics.
• /consensus/hotstuff/pacemaker contains the implementation of Flow's basic PaceMaker, as described above.
• /consensus/hotstuff/persister for performance reasons, the implementation maintains the consensus state largely in-memory. The persister stores the last entered view and the view of the latest voted block persistenlty on disk. This allows recovery after a crash without the risk of equivocation.
• /consensus/hotstuff/runner helper code for starting and shutting down the HotStuff logic safely in a multithreaded environment.
• /consensus/hotstuff/validator holds the logic for validating the HotStuff-relevant aspects of blocks, QCs, and votes
• /consensus/hotstuff/verification contains integration of Flow's cryptographic primitives (signing and signature verification)
• /consensus/hotstuff/voteaggregator caches votes on a per-block basis and builds a QC if enough votes have been accumulated.
• /consensus/hotstuff/voter tracks the view of the latest vote and determines whether or not to vote for a block

Pending Extensions

• BLS Aggregation of the StakingSignatures
• include Epochs
• upgrade PaceMaker to include Timeout Certificates
• refactor crypto integration (code in verification and dependent modules) for better auditability

Telemetry

The HotStuff state machine exposes some details about its internal progress as notification through the hotstuff.Consumer. The following figure depicts at which points notifications are emitted.

We have implemented a telemetry system (hotstuff.notifications.TelemetryConsumer) which implements the Consumer interface. The TelemetryConsumer tracks all events as belonging together that were emitted during a path through the state machine. Each path through the state machine is identified by a unique id. Generally, the TelemetryConsumer could export the collected data to a variety of backends. For now, we export the data to a logger.

Documentation

Index

• type BlockProducer
• type BlockSignatureData
• type BlockSignerDecoder
• type CommunicatorConsumer
• type Consumer
• type DKG
• type DynamicCommittee
• type EventHandler
• type EventLoop
• type FinalizationConsumer
• type FollowerLogic
• type FollowerLoop
  • func NewFollowerLoop(log zerolog.Logger, followerLogic FollowerLogic) (*FollowerLoop, error)
  • func (fl *FollowerLoop) SubmitProposal(proposal *model.Proposal)
• type Forks
• type LivenessData
• type OnQCCreated
• type PaceMaker
• type Packer
• type PartialTcCreated
• type Persister
• type QCCreatedConsumer
• type RandomBeaconInspector
• type RandomBeaconReconstructor
• type Replicas
• type SafetyData
• type SafetyRules
• type Signer
• type TimeoutAggregator
• type TimeoutCollector
• type TimeoutCollectorConsumer
• type TimeoutCollectorFactory
• type TimeoutCollectors
• type TimeoutProcessor
• type TimeoutProcessorFactory
• type TimeoutSignatureAggregator
• type TimeoutSignerInfo
• type Validator
• type Verifier
• type VerifyingVoteProcessor
• type VoteAggregator
• type VoteCollector
• type VoteCollectorStatus
  • func (ps VoteCollectorStatus) String() string
• type VoteCollectors
• type VoteConsumer
• type VoteProcessor
• type VoteProcessorFactory
• type WeightedSignatureAggregator
• type Workerpool
• type Workers

Types

type BlockProducer

type BlockProducer interface {
	// MakeBlockProposal builds a new HotStuff block proposal using the given view,
	// the given quorum certificate for its parent and [optionally] a timeout certificate for last view(could be nil).
	// No errors are expected during normal operation.
	MakeBlockProposal(view uint64, qc *flow.QuorumCertificate, lastViewTC *flow.TimeoutCertificate) (*flow.Header, error)
}

BlockProducer builds a new block proposal by building a new block payload with the builder module, and uses VoteCollectorFactory to create a disposable VoteCollector for producing the proposal vote. BlockProducer assembles the new block proposal using the block payload, block header and the proposal vote.

type BlockSignatureData

type BlockSignatureData struct {
	StakingSigners               flow.IdentifierList
	RandomBeaconSigners          flow.IdentifierList
	AggregatedStakingSig         []byte // if BLS is used, this is equivalent to crypto.Signature
	AggregatedRandomBeaconSig    []byte // if BLS is used, this is equivalent to crypto.Signature
	ReconstructedRandomBeaconSig crypto.Signature
}

BlockSignatureData is an intermediate struct for Packer to pack the aggregated signature data into raw bytes or unpack from raw bytes.

type BlockSignerDecoder

type BlockSignerDecoder interface {
	// DecodeSignerIDs decodes the signer indices from the given block header into full node IDs.
	// Note: A block header contains a quorum certificate for its parent, which proves that the
	// consensus committee has reached agreement on validity of parent block. Consequently, the
	// returned IdentifierList contains the consensus participants that signed the parent block.
	// Expected Error returns during normal operations:
	//   - model.ErrViewForUnknownEpoch if the given block's parent is within an unknown epoch
	//   - signature.InvalidSignerIndicesError if signer indices included in the header do
	//     not encode a valid subset of the consensus committee
	DecodeSignerIDs(header *flow.Header) (flow.IdentifierList, error)
}

BlockSignerDecoder defines how to convert the ParentSignerIndices field within a particular block header to the identifiers of the nodes which signed the block.

type CommunicatorConsumer

type CommunicatorConsumer interface {
	// OnOwnVote notifies about intent to send a vote for the given parameters to the specified recipient.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnOwnVote(blockID flow.Identifier, view uint64, sigData []byte, recipientID flow.Identifier)

	// OnOwnTimeout notifies about intent to broadcast the given timeout object(TO) to all actors of the consensus process.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnOwnTimeout(timeout *model.TimeoutObject)

	// OnOwnProposal notifies about intent to broadcast the given block proposal to all actors of
	// the consensus process.
	// delay is to hold the proposal before broadcasting it. Useful to control the block production rate.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnOwnProposal(proposal *flow.Header, targetPublicationTime time.Time)
}

CommunicatorConsumer consumes outbound notifications produced by HotStuff and it's components. Notifications allow the HotStuff core algorithm to communicate with the other actors of the consensus process. Implementations must:

• be concurrency safe
• be non-blocking
• handle repetition of the same events (with some processing overhead).

type Consumer

type Consumer interface {
	FinalizationConsumer
	CommunicatorConsumer

	// OnEventProcessed notifications are produced by the EventHandler when it is done processing
	// and hands control back to the EventLoop to wait for the next event.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnEventProcessed()

	// OnStart notifications are produced by the EventHandler when it starts blocks recovery and
	// prepares for handling incoming events from EventLoop.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnStart(currentView uint64)

	// OnReceiveProposal notifications are produced by the EventHandler when it starts processing a block.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnReceiveProposal(currentView uint64, proposal *model.Proposal)

	// OnReceiveQc notifications are produced by the EventHandler when it starts processing a
	// QuorumCertificate [QC] constructed by the node's internal vote aggregator.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnReceiveQc(currentView uint64, qc *flow.QuorumCertificate)

	// OnReceiveTc notifications are produced by the EventHandler when it starts processing a
	// TimeoutCertificate [TC]  constructed by the node's internal timeout aggregator.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnReceiveTc(currentView uint64, tc *flow.TimeoutCertificate)

	// OnPartialTc notifications are produced by the EventHandler when it starts processing partial TC
	// constructed by local timeout aggregator.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnPartialTc(currentView uint64, partialTc *PartialTcCreated)

	// OnLocalTimeout notifications are produced by the EventHandler when it reacts to expiry of round duration timer.
	// Such a notification indicates that the PaceMaker's timeout was processed by the system.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnLocalTimeout(currentView uint64)

	// OnViewChange notifications are produced by PaceMaker when it transitions to a new view
	// based on processing a QC or TC. The arguments specify the oldView (first argument),
	// and the newView to which the PaceMaker transitioned (second argument).
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnViewChange(oldView, newView uint64)

	// OnQcTriggeredViewChange notifications are produced by PaceMaker when it moves to a new view
	// based on processing a QC. The arguments specify the qc (first argument), which triggered
	// the view change, and the newView to which the PaceMaker transitioned (second argument).
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnQcTriggeredViewChange(oldView uint64, newView uint64, qc *flow.QuorumCertificate)

	// OnTcTriggeredViewChange notifications are produced by PaceMaker when it moves to a new view
	// based on processing a TC. The arguments specify the tc (first argument), which triggered
	// the view change, and the newView to which the PaceMaker transitioned (second argument).
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnTcTriggeredViewChange(oldView uint64, newView uint64, tc *flow.TimeoutCertificate)

	// OnStartingTimeout notifications are produced by PaceMaker. Such a notification indicates that the
	// PaceMaker is now waiting for the system to (receive and) process blocks or votes.
	// The specific timeout type is contained in the TimerInfo.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnStartingTimeout(model.TimerInfo)

	// OnVoteProcessed notifications are produced by the Vote Aggregation logic, each time
	// we successfully ingest a valid vote.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnVoteProcessed(vote *model.Vote)

	// OnTimeoutProcessed notifications are produced by the Timeout Aggregation logic,
	// each time we successfully ingest a valid timeout.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnTimeoutProcessed(timeout *model.TimeoutObject)

	// OnCurrentViewDetails notifications are produced by the EventHandler during the course of a view with auxiliary information.
	// These notifications are generally not produced for all views (for example skipped views).
	// These notifications are guaranteed to be produced for all views we enter after fully processing a message.
	// Example 1:
	//   - We are in view 8. We process a QC with view 10, causing us to enter view 11.
	//   - Then this notification will be produced for view 11.
	// Example 2:
	//   - We are in view 8. We process a proposal with view 10, which contains a TC for view 9 and TC.NewestQC for view 8.
	//   - The QC would allow us to enter view 9 and the TC would allow us to enter view 10,
	//     so after fully processing the message we are in view 10.
	//   - Then this notification will be produced for view 10, but not view 9
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnCurrentViewDetails(currentView, finalizedView uint64, currentLeader flow.Identifier)

	// OnDoubleVotingDetected notifications are produced by the Vote Aggregation logic
	// whenever a double voting (same voter voting for different blocks at the same view) was detected.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnDoubleVotingDetected(*model.Vote, *model.Vote)

	// OnInvalidVoteDetected notifications are produced by the Vote Aggregation logic
	// whenever an invalid vote was detected.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnInvalidVoteDetected(err model.InvalidVoteError)

	// OnVoteForInvalidBlockDetected notifications are produced by the Vote Aggregation logic
	// whenever vote for invalid proposal was detected.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnVoteForInvalidBlockDetected(vote *model.Vote, invalidProposal *model.Proposal)

	// OnDoubleTimeoutDetected notifications are produced by the Timeout Aggregation logic
	// whenever a double timeout (same replica producing two different timeouts at the same view) was detected.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnDoubleTimeoutDetected(*model.TimeoutObject, *model.TimeoutObject)

	// OnInvalidTimeoutDetected notifications are produced by the Timeout Aggregation logic
	// whenever an invalid timeout was detected.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnInvalidTimeoutDetected(err model.InvalidTimeoutError)
}

Consumer consumes outbound notifications produced by HotStuff and its components. Notifications are consensus-internal state changes which are potentially relevant to the larger node in which HotStuff is running. The notifications are emitted in the order in which the HotStuff algorithm makes the respective steps.

Implementations must:

• be concurrency safe
• be non-blocking
• handle repetition of the same events (with some processing overhead).

type DKG

type DKG interface {
	protocol.DKG
}

type DynamicCommittee

type DynamicCommittee interface {
	Replicas

	// IdentitiesByBlock returns a list of the legitimate HotStuff participants for the given block.
	// The list of participants is filtered by the provided selector.
	// The returned list of HotStuff participants:
	//   * contains nodes that are allowed to submit proposals, votes, and timeouts
	//     (un-ejected, non-zero weight at current block)
	//   * is ordered in the canonical order
	//   * contains no duplicates.
	// The list of all legitimate HotStuff participants for the given epoch can be obtained by using `filter.Any`
	//
	// TODO - do we need this, if we are only checking a single proposer ID?
	IdentitiesByBlock(blockID flow.Identifier) (flow.IdentityList, error)

	// IdentityByBlock returns the full Identity for specified HotStuff participant.
	// The node must be a legitimate HotStuff participant with NON-ZERO WEIGHT at the specified block.
	// ERROR conditions:
	//  * model.InvalidSignerError if participantID does NOT correspond to an authorized HotStuff participant at the specified block.
	IdentityByBlock(blockID flow.Identifier, participantID flow.Identifier) (*flow.Identity, error)
}

DynamicCommittee extends Replicas to provide the consensus committee for the purposes of validating proposals. The proposer committee reflects block-to-block changes in the identity table to support immediately rejecting proposals from nodes after they are ejected. For validating proposals, we use *ByBlock methods.

Since the proposer committee can change at any block: * we query by block ID * we must have incorporated the full block ancestry prior to validating messages

type EventHandler

type EventHandler interface {

	// OnReceiveQc processes a valid qc constructed by internal vote aggregator or discovered in TimeoutObject.
	// All inputs should be validated before feeding into this function. Assuming trusted data.
	// No errors are expected during normal operation.
	OnReceiveQc(qc *flow.QuorumCertificate) error

	// OnReceiveTc processes a valid tc constructed by internal timeout aggregator, discovered in TimeoutObject or
	// broadcast over the network.
	// All inputs should be validated before feeding into this function. Assuming trusted data.
	// No errors are expected during normal operation.
	OnReceiveTc(tc *flow.TimeoutCertificate) error

	// OnReceiveProposal processes a block proposal received from another HotStuff
	// consensus participant.
	// All inputs should be validated before feeding into this function. Assuming trusted data.
	// No errors are expected during normal operation.
	OnReceiveProposal(proposal *model.Proposal) error

	// OnLocalTimeout handles a local timeout event by creating a model.TimeoutObject and broadcasting it.
	// No errors are expected during normal operation.
	OnLocalTimeout() error

	// OnPartialTcCreated handles notification produces by the internal timeout aggregator. If the notification is for the current view,
	// a corresponding model.TimeoutObject is broadcast to the consensus committee.
	// No errors are expected during normal operation.
	OnPartialTcCreated(partialTC *PartialTcCreated) error

	// TimeoutChannel returns a channel that sends a signal on timeout.
	TimeoutChannel() <-chan time.Time

	// Start starts the event handler.
	// No errors are expected during normal operation.
	// CAUTION: EventHandler is not concurrency safe. The Start method must
	// be executed by the same goroutine that also calls the other business logic
	// methods, or concurrency safety has to be implemented externally.
	Start(ctx context.Context) error
}

EventHandler runs a state machine to process proposals, QC and local timeouts. Not concurrency safe.

type EventLoop

type EventLoop interface {
	module.HotStuff
	TimeoutCollectorConsumer
	QCCreatedConsumer
}

EventLoop performs buffer and processing of incoming proposals and QCs.

type FinalizationConsumer

type FinalizationConsumer interface {

	// OnBlockIncorporated notifications are produced by the Finalization Logic
	// whenever a block is incorporated into the consensus state.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnBlockIncorporated(*model.Block)

	// OnFinalizedBlock notifications are produced by the Finalization Logic whenever
	// a block has been finalized. They are emitted in the order the blocks are finalized.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnFinalizedBlock(*model.Block)

	// OnDoubleProposeDetected notifications are produced by the Finalization Logic
	// whenever a double block proposal (equivocation) was detected.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnDoubleProposeDetected(*model.Block, *model.Block)
}

FinalizationConsumer consumes outbound notifications produced by the finalization logic. Notifications represent finalization-specific state changes which are potentially relevant to the larger node. The notifications are emitted in the order in which the finalization algorithm makes the respective steps.

Implementations must:

• be concurrency safe
• be non-blocking
• handle repetition of the same events (with some processing overhead).

type FollowerLogic

type FollowerLogic interface {
	// FinalizedBlock returns the latest finalized block
	FinalizedBlock() *model.Block

	// AddBlock processes a block proposal
	AddBlock(proposal *model.Proposal) error
}

FollowerLogic runs a state machine to process proposals

type FollowerLoop

type FollowerLoop struct {
	*component.ComponentManager
	// contains filtered or unexported fields
}

FollowerLoop implements interface module.HotStuffFollower. FollowerLoop buffers all incoming events to the hotstuff FollowerLogic, and feeds FollowerLogic one event at a time using a worker thread. Concurrency safe.

func NewFollowerLoop

func NewFollowerLoop(log zerolog.Logger, followerLogic FollowerLogic) (*FollowerLoop, error)

NewFollowerLoop creates an instance of EventLoop

func (*FollowerLoop) SubmitProposal

func (fl *FollowerLoop) SubmitProposal(proposal *model.Proposal)

SubmitProposal feeds a new block proposal (header) into the FollowerLoop. This method blocks until the proposal is accepted to the event queue.

Block proposals must be submitted in order, i.e. a proposal's parent must have been previously processed by the FollowerLoop.

type Forks

type Forks interface {

	// GetProposalsForView returns all BlockProposals at the given view number.
	GetProposalsForView(view uint64) []*model.Proposal

	// GetProposal returns (BlockProposal, true) if the block with the specified
	// id was found (nil, false) otherwise.
	GetProposal(id flow.Identifier) (*model.Proposal, bool)

	// FinalizedView returns the largest view number where a finalized block is known
	FinalizedView() uint64

	// FinalizedBlock returns the finalized block with the largest view number
	FinalizedBlock() *model.Block

	// NewestView returns the largest view number of all proposals that were added to Forks.
	NewestView() uint64

	// AddProposal adds the block proposal to Forks. This might cause an update of the finalized block
	// and pruning of older blocks.
	// Handles duplicated addition of blocks (at the potential cost of additional computation time).
	// PREREQUISITE:
	// Forks must be able to connect `proposal` to its latest finalized block
	// (without missing interim ancestors). Otherwise, an exception is raised.
	// Expected errors during normal operations:
	//  * model.ByzantineThresholdExceededError - new block results in conflicting finalized blocks
	AddProposal(proposal *model.Proposal) error
}

Forks maintains an in-memory data-structure of all proposals whose view-number is larger or equal to the latest finalized block. The latest finalized block is defined as the finalized block with the largest view number. When adding blocks, Forks automatically updates its internal state (including finalized blocks). Furthermore, blocks whose view number is smaller than the latest finalized block are pruned automatically.

PREREQUISITES: Forks expects that only blocks are added that can be connected to its latest finalized block (without missing interim ancestors). If this condition is violated, Forks will raise an error and ignore the block.

type LivenessData

type LivenessData struct {
	// CurrentView is the currently active view tracked by the PaceMaker. It is updated
	// whenever the PaceMaker sees evidence (QC or TC) for advancing to next view.
	CurrentView uint64
	// NewestQC is the newest QC (by view) observed by the PaceMaker. The QC can be observed on its own or as a part of TC.
	NewestQC *flow.QuorumCertificate
	// LastViewTC is the TC for the prior view (CurrentView-1), if this view timed out. If the previous round
	// ended with a QC, this QC is stored in NewestQC and LastViewTC is nil.
	LastViewTC *flow.TimeoutCertificate
}

type OnQCCreated

type OnQCCreated func(*flow.QuorumCertificate)

OnQCCreated is a callback which will be used by VoteCollector to submit a QC when it's able to create it

type PaceMaker

type PaceMaker interface {

	// CurView returns the current view.
	CurView() uint64

	// NewestQC returns QC with the highest view discovered by PaceMaker.
	NewestQC() *flow.QuorumCertificate

	// LastViewTC returns TC for last view, this could be nil if previous round
	// has entered with a QC.
	LastViewTC() *flow.TimeoutCertificate

	// ProcessQC will check if the given QC will allow PaceMaker to fast-forward to QC.view+1.
	// If PaceMaker incremented the current View, a NewViewEvent will be returned.
	// No errors are expected during normal operation.
	ProcessQC(qc *flow.QuorumCertificate) (*model.NewViewEvent, error)

	// ProcessTC will check if the given TC will allow PaceMaker to fast-forward to TC.view+1.
	// If PaceMaker incremented the current View, a NewViewEvent will be returned.
	// A nil TC is an expected valid input.
	// No errors are expected during normal operation.
	ProcessTC(tc *flow.TimeoutCertificate) (*model.NewViewEvent, error)

	// TimeoutChannel returns the timeout channel for the CURRENTLY ACTIVE timeout.
	// Each time the pacemaker starts a new timeout, this channel is replaced.
	TimeoutChannel() <-chan time.Time

	// Start starts the PaceMaker (i.e. the timeout for the configured starting value for view).
	// CAUTION: EventHandler is not concurrency safe. The Start method must
	// be executed by the same goroutine that also calls the other business logic
	// methods, or concurrency safety has to be implemented externally.
	Start(ctx context.Context)

	// BlockRateDelay returns the minimal wait time for broadcasting a proposal, measured from
	// the point in time when the primary (locally) enters the respective view.
	BlockRateDelay() time.Duration
}

PaceMaker for HotStuff. The component is passive in that it only reacts to method calls. The PaceMaker does not perform state transitions on its own. Timeouts are emitted through channels. Each timeout has its own dedicated channel, which is garbage collected after the respective state has been passed. It is the EventHandler's responsibility to pick up timeouts from the currently active TimeoutChannel process them first and subsequently inform the PaceMaker about processing the timeout. Specifically, the intended usage pattern for the TimeoutChannels is as follows:

• Each time the PaceMaker starts a new timeout, it created a new TimeoutChannel

• The channel for the CURRENTLY ACTIVE timeout is returned by PaceMaker.TimeoutChannel()

• Each time the EventHandler processes an event, the EventHandler might call into PaceMaker potentially resulting in a state transition and the PaceMaker starting a new timeout

• Hence, after processing any event, EventHandler should retrieve the current TimeoutChannel from the PaceMaker.

For Example:

for {
	timeoutChannel := el.eventHandler.TimeoutChannel()
	select {
	   case <-timeoutChannel:
	    	el.eventHandler.OnLocalTimeout()
	   case <other events>
	}
}

Not concurrency safe.

type Packer

type Packer interface {
	// Pack serializes the provided BlockSignatureData into a precursor format of a QC.
	// view is the view of the block that the aggregated signature is for.
	// sig is the aggregated signature data.
	// Expected error returns during normal operations:
	//  * none; all errors are symptoms of inconsistent input data or corrupted internal state.
	Pack(view uint64, sig *BlockSignatureData) (signerIndices []byte, sigData []byte, err error)

	// Unpack de-serializes the provided signature data.
	// sig is the aggregated signature data
	// It returns:
	//  - (sigData, nil) if successfully unpacked the signature data
	//  - (nil, model.InvalidFormatError) if failed to unpack the signature data
	Unpack(signerIdentities flow.IdentityList, sigData []byte) (*BlockSignatureData, error)
}

Packer packs aggregated signature data into raw bytes to be used in block header.

type PartialTcCreated

type PartialTcCreated struct {
	View       uint64
	NewestQC   *flow.QuorumCertificate
	LastViewTC *flow.TimeoutCertificate
}

PartialTcCreated represents a notification emitted by the TimeoutProcessor component, whenever it has collected TimeoutObjects from a superminority of consensus participants for a specific view. Along with the view, it reports the newest QC and TC (for previous view) discovered during timeout collection. Per convention, the newest QC is never nil, while the TC for the previous view might be nil.

type Persister

type Persister interface {
	// GetSafetyData will retrieve last persisted safety data.
	// During normal operations, no errors are expected.
	GetSafetyData() (*SafetyData, error)

	// PutSafetyData persists the last safety data.
	// This method blocks until `safetyData` was successfully persisted.
	// During normal operations, no errors are expected.
	PutSafetyData(safetyData *SafetyData) error

	// GetLivenessData will retrieve last persisted liveness data.
	// During normal operations, no errors are expected.
	GetLivenessData() (*LivenessData, error)

	// PutLivenessData persists the last liveness data.
	// This method blocks until `safetyData` was successfully persisted.
	// During normal operations, no errors are expected.
	PutLivenessData(livenessData *LivenessData) error
}

Persister is responsible for persisting state we need to bootstrap after a restart or crash.

type QCCreatedConsumer

type QCCreatedConsumer interface {
	// OnQcConstructedFromVotes notifications are produced by the VoteAggregator
	// component, whenever it constructs a QC from votes.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnQcConstructedFromVotes(*flow.QuorumCertificate)
}

QCCreatedConsumer consumes outbound notifications produced by HotStuff and its components. Notifications are consensus-internal state changes which are potentially relevant to the larger node in which HotStuff is running. The notifications are emitted in the order in which the HotStuff algorithm makes the respective steps.

Implementations must:

• be concurrency safe
• be non-blocking
• handle repetition of the same events (with some processing overhead).

type RandomBeaconInspector

type RandomBeaconInspector interface {
	// Verify verifies the signature share under the signer's public key and the message agreed upon.
	// The function is thread-safe and wait-free (i.e. allowing arbitrary many routines to
	// execute the business logic, without interfering with each other).
	// It allows concurrent verification of the given signature.
	// Returns :
	//  - model.InvalidSignerError if signerIndex is invalid
	//  - model.ErrInvalidSignature if signerIndex is valid but signature is cryptographically invalid
	//  - other error if there is an unexpected exception.
	Verify(signerIndex int, share crypto.Signature) error

	// TrustedAdd adds a share to the internal signature shares store.
	// There is no pre-check of the signature's validity _before_ adding it.
	// It is the caller's responsibility to make sure the signature was previously verified.
	// Nevertheless, the implementation guarantees safety (only correct threshold signatures
	// are returned) through a post-check (verifying the threshold signature
	// _after_ reconstruction before returning it).
	// The function is thread-safe but locks its internal state, thereby permitting only
	// one routine at a time to add a signature.
	// Returns:
	//  - (true, nil) if the signature has been added, and enough shares have been collected.
	//  - (false, nil) if the signature has been added, but not enough shares were collected.
	//  - (false, error) if there is any exception adding the signature share.
	//      - model.InvalidSignerError if signerIndex is invalid (out of the valid range)
	//  	- model.DuplicatedSignerError if the signer has been already added
	//      - other error if there is an unexpected exception.
	TrustedAdd(signerIndex int, share crypto.Signature) (enoughshares bool, exception error)

	// EnoughShares indicates whether enough shares have been accumulated in order to reconstruct
	// a group signature. The function is thread-safe.
	EnoughShares() bool

	// Reconstruct reconstructs the group signature. The function is thread-safe but locks
	// its internal state, thereby permitting only one routine at a time.
	//
	// Returns:
	// - (signature, nil) if no error occurred
	// - (nil, model.InsufficientSignaturesError) if not enough shares were collected
	// - (nil, model.InvalidSignatureIncluded) if at least one collected share does not serialize to a valid BLS signature,
	//    or if the constructed signature failed to verify against the group public key and stored message. This post-verification
	//    is required  for safety, as `TrustedAdd` allows adding invalid signatures.
	// - (nil, error) for any other unexpected error.
	Reconstruct() (crypto.Signature, error)
}

RandomBeaconInspector encapsulates all methods needed by a Hotstuff leader to validate the beacon votes and reconstruct a beacon signature. The random beacon methods are based on a threshold signature scheme.

type RandomBeaconReconstructor

type RandomBeaconReconstructor interface {
	// Verify verifies the signature share under the signer's public key and the message agreed upon.
	// The function is thread-safe and wait-free (i.e. allowing arbitrary many routines to
	// execute the business logic, without interfering with each other).
	// It allows concurrent verification of the given signature.
	// Returns :
	//  - model.InvalidSignerError if signerIndex is invalid
	//  - model.ErrInvalidSignature if signerID is valid but signature is cryptographically invalid
	//  - other error if there is an unexpected exception.
	Verify(signerID flow.Identifier, sig crypto.Signature) error

	// TrustedAdd adds a share to the internal signature shares store.
	// There is no pre-check of the signature's validity _before_ adding it.
	// It is the caller's responsibility to make sure the signature was previously verified.
	// Nevertheless, the implementation guarantees safety (only correct threshold signatures
	// are returned) through a post-check (verifying the threshold signature
	// _after_ reconstruction before returning it).
	// The function is thread-safe but locks its internal state, thereby permitting only
	// one routine at a time to add a signature.
	// Returns:
	//  - (true, nil) if the signature has been added, and enough shares have been collected.
	//  - (false, nil) if the signature has been added, but not enough shares were collected.
	//  - (false, error) if there is any exception adding the signature share.
	//      - model.InvalidSignerError if signerIndex is invalid (out of the valid range)
	//  	- model.DuplicatedSignerError if the signer has been already added
	//      - other error if there is an unexpected exception.
	TrustedAdd(signerID flow.Identifier, sig crypto.Signature) (EnoughShares bool, err error)

	// EnoughShares indicates whether enough shares have been accumulated in order to reconstruct
	// a group signature. The function is thread-safe.
	EnoughShares() bool

	// Reconstruct reconstructs the group signature. The function is thread-safe but locks
	// its internal state, thereby permitting only one routine at a time.
	//
	// Returns:
	// - (signature, nil) if no error occurred
	// - (nil, model.InsufficientSignaturesError) if not enough shares were collected
	// - (nil, model.InvalidSignatureIncluded) if at least one collected share does not serialize to a valid BLS signature,
	//    or if the constructed signature failed to verify against the group public key and stored message. This post-verification
	//    is required  for safety, as `TrustedAdd` allows adding invalid signatures.
	// - (nil, error) for any other unexpected error.
	Reconstruct() (crypto.Signature, error)
}

RandomBeaconReconstructor encapsulates all methods needed by a Hotstuff leader to validate the beacon votes and reconstruct a beacon signature. The random beacon methods are based on a threshold signature scheme.

type Replicas

type Replicas interface {

	// LeaderForView returns the identity of the leader for a given view.
	// CAUTION: per liveness requirement of HotStuff, the leader must be fork-independent.
	//          Therefore, a node retains its proposer view slots even if it is slashed.
	//          Its proposal is simply considered invalid, as it is not from a legitimate participant.
	// Returns the following expected errors for invalid inputs:
	//   * model.ErrViewForUnknownEpoch if no epoch containing the given view is known
	LeaderForView(view uint64) (flow.Identifier, error)

	// QuorumThresholdForView returns the minimum total weight for a supermajority
	// at the given view. This weight threshold is computed using the total weight
	// of the initial committee and is static over the course of an epoch.
	// Returns the following expected errors for invalid inputs:
	//   * model.ErrViewForUnknownEpoch if no epoch containing the given view is known
	QuorumThresholdForView(view uint64) (uint64, error)

	// TimeoutThresholdForView returns the minimum total weight of observed timeout objects
	// required to safely timeout for the given view. This weight threshold is computed
	// using the total weight of the initial committee and is static over the course of
	// an epoch.
	// Returns the following expected errors for invalid inputs:
	//   * model.ErrViewForUnknownEpoch if no epoch containing the given view is known
	TimeoutThresholdForView(view uint64) (uint64, error)

	// Self returns our own node identifier.
	// TODO: ultimately, the own identity of the node is necessary for signing.
	//       Ideally, we would move the method for checking whether an Identifier refers to this node to the signer.
	//       This would require some refactoring of EventHandler (postponed to later)
	Self() flow.Identifier

	// DKG returns the DKG info for epoch given by the input view.
	// Returns the following expected errors for invalid inputs:
	//   * model.ErrViewForUnknownEpoch if no epoch containing the given view is known
	DKG(view uint64) (DKG, error)

	// IdentitiesByEpoch returns a list of the legitimate HotStuff participants for the epoch
	// given by the input view. The list of participants is filtered by the provided selector.
	// The returned list of HotStuff participants:
	//   * contains nodes that are allowed to submit votes or timeouts within the given epoch
	//     (un-ejected, non-zero weight at the beginning of the epoch)
	//   * is ordered in the canonical order
	//   * contains no duplicates.
	// The list of all legitimate HotStuff participants for the given epoch can be obtained by using `filter.Any`
	//
	// CAUTION: DO NOT use this method for validating block proposals.
	//
	// Returns the following expected errors for invalid inputs:
	//   * model.ErrViewForUnknownEpoch if no epoch containing the given view is known
	//
	// TODO: should return identity skeleton https://github.com/dapperlabs/flow-go/issues/6232
	IdentitiesByEpoch(view uint64) (flow.IdentityList, error)

	// IdentityByEpoch returns the full Identity for specified HotStuff participant.
	// The node must be a legitimate HotStuff participant with NON-ZERO WEIGHT at the specified block.
	// ERROR conditions:
	//  * model.InvalidSignerError if participantID does NOT correspond to an authorized HotStuff participant at the specified block.
	//
	// Returns the following expected errors for invalid inputs:
	//   * model.ErrViewForUnknownEpoch if no epoch containing the given view is known
	//
	// TODO: should return identity skeleton https://github.com/dapperlabs/flow-go/issues/6232
	IdentityByEpoch(view uint64, participantID flow.Identifier) (*flow.Identity, error)
}

Replicas defines the consensus committee for the purposes of validating votes, timeouts, quorum certificates, and timeout certificates. Any consensus committee member who was authorized to contribute to consensus AT THE BEGINNING of the epoch may produce valid votes and timeouts for the entire epoch, even if they are later ejected. So for validating votes/timeouts we use *ByEpoch methods.

Since the voter committee is considered static over an epoch: * we can query identities by view * we don't need the full block ancestry prior to validating messages

type SafetyData

type SafetyData struct {
	// LockedOneChainView is the head block's view of the newest 1-chain this replica has voted for.
	// The 1-chain can be indirect.
	//     <·· <QC>[B0] <- <QC_B0>[B1] <- [my vote for B1]
	// In the depicted scenario, the replica voted for block B1, which forms a (potentially indirect)
	// 1-chain on top of B0. The replica updated LockedOneChainView to the max of the current value and
	// QC_B0.View = B0.View. Thereby, the safety module guarantees that the replica will not sign
	// a TimeoutObject that would allow a malicious leader to fork below the latest finalized block.
	LockedOneChainView uint64
	// HighestAcknowledgedView is the highest view where we have voted or triggered a timeout
	HighestAcknowledgedView uint64
	// LastTimeout is the last timeout that was produced by this node (may be nil if no timeout occurred yet)
	LastTimeout *model.TimeoutObject
}

type SafetyRules

type SafetyRules interface {
	// ProduceVote takes a block proposal and current view, and decides whether to vote for the block.
	// Voting is deterministic meaning voting for same proposal will always result in the same vote.
	// Returns:
	//  * (vote, nil): On the _first_ block for the current view that is safe to vote for.
	//    Subsequently, voter does _not_ vote for any _other_  block with the same (or lower) view.
	//    SafetyRules internally caches and persists its latest vote. As long as the SafetyRules' internal
	//    state remains unchanged, ProduceVote will return its cached for identical inputs.
	//  * (nil, model.NoVoteError): If the safety module decides that it is not safe to vote for the given block.
	//    This is a sentinel error and _expected_ during normal operation.
	// All other errors are unexpected and potential symptoms of uncovered edge cases or corrupted internal state (fatal).
	ProduceVote(proposal *model.Proposal, curView uint64) (*model.Vote, error)
	// ProduceTimeout takes current view, highest locally known QC and TC (optional, must be nil if and
	// only if QC is for previous view) and decides whether to produce timeout for current view.
	// Returns:
	//  * (timeout, nil): It is safe to timeout for current view using newestQC and lastViewTC.
	//  * (nil, model.NoTimeoutError): If replica is not part of the authorized consensus committee (anymore) and
	//    therefore is not authorized to produce a valid timeout object. This sentinel error is _expected_ during
	//    normal operation, e.g. during the grace-period after Epoch switchover or after the replica self-ejected.
	// All other errors are unexpected and potential symptoms of uncovered edge cases or corrupted internal state (fatal).
	ProduceTimeout(curView uint64, newestQC *flow.QuorumCertificate, lastViewTC *flow.TimeoutCertificate) (*model.TimeoutObject, error)
}

SafetyRules enforces all consensus rules that guarantee safety. It produces votes for the given blocks or TimeoutObject for the given views, only if all safety rules are satisfied.

type Signer

type Signer interface {
	// CreateProposal creates a proposal for the given block. No error returns
	// are expected during normal operations (incl. presence of byz. actors).
	CreateProposal(block *model.Block) (*model.Proposal, error)

	// CreateVote creates a vote for the given block. No error returns are
	// expected during normal operations (incl. presence of byz. actors).
	CreateVote(block *model.Block) (*model.Vote, error)

	// CreateTimeout creates a timeout for given view. No errors return are
	// expected during normal operations(incl presence of byz. actors).
	CreateTimeout(curView uint64, newestQC *flow.QuorumCertificate, lastViewTC *flow.TimeoutCertificate) (*model.TimeoutObject, error)
}

Signer is responsible for creating votes, proposals for a given block.

type TimeoutAggregator

type TimeoutAggregator interface {
	module.ReadyDoneAware
	module.Startable

	// AddTimeout verifies and aggregates a timeout object.
	// This method can be called concurrently, timeouts will be queued and processed asynchronously.
	AddTimeout(timeoutObject *model.TimeoutObject)

	// PruneUpToView deletes all `TimeoutCollector`s _below_ to the given view, as well as
	// related indices. We only retain and process `TimeoutCollector`s, whose view is equal or larger
	// than `lowestRetainedView`. If `lowestRetainedView` is smaller than the
	// previous value, the previous value is kept and the method call is a NoOp.
	// This value should be set to the latest active view maintained by `Pacemaker`.
	PruneUpToView(lowestRetainedView uint64)
}

TimeoutAggregator verifies and aggregates timeout objects to build timeout certificates [TCs]. When enough timeout objects are collected, it builds a TC and sends it to the EventLoop TimeoutAggregator also detects protocol violation, including invalid timeouts, double timeout, etc and notifies a HotStuff consumer for slashing.

type TimeoutCollector

type TimeoutCollector interface {
	// AddTimeout adds a Timeout Object [TO] to the collector.
	// When TOs from strictly more than 1/3 of consensus participants (measured by weight)
	// were collected, the callback for partial TC will be triggered.
	// After collecting TOs from a supermajority, a TC will be created and passed to the EventLoop.
	// Expected error returns during normal operations:
	// * timeoutcollector.ErrTimeoutForIncompatibleView - submitted timeout for incompatible view
	// All other exceptions are symptoms of potential state corruption.
	AddTimeout(timeoutObject *model.TimeoutObject) error

	// View returns the view that this instance is collecting timeouts for.
	// This method is useful when adding the newly created timeout collector to timeout collectors map.
	View() uint64
}

TimeoutCollector collects all timeout objects for a specified view. On the happy path, it generates a TimeoutCertificate when enough timeouts have been collected. The TimeoutCollector is a higher-level structure that orchestrates deduplication, caching and processing of timeouts, delegating those tasks to underlying modules (such as TimeoutProcessor). Implementations of TimeoutCollector must be concurrency safe.

type TimeoutCollectorConsumer

type TimeoutCollectorConsumer interface {
	// OnTcConstructedFromTimeouts notifications are produced by the TimeoutProcessor
	// component, whenever it constructs a TC based on TimeoutObjects from a
	// supermajority of consensus participants.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnTcConstructedFromTimeouts(certificate *flow.TimeoutCertificate)

	// OnPartialTcCreated notifications are produced by the TimeoutProcessor
	// component, whenever it collected TimeoutObjects from a superminority
	// of consensus participants for a specific view. Along with the view, it
	// reports the newest QC and TC (for previous view) discovered in process of
	// timeout collection. Per convention, the newest QC is never nil, while
	// the TC for the previous view might be nil.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnPartialTcCreated(view uint64, newestQC *flow.QuorumCertificate, lastViewTC *flow.TimeoutCertificate)

	// OnNewQcDiscovered notifications are produced by the TimeoutCollector
	// component, whenever it discovers new QC included in timeout object.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnNewQcDiscovered(certificate *flow.QuorumCertificate)

	// OnNewTcDiscovered notifications are produced by the TimeoutCollector
	// component, whenever it discovers new TC included in timeout object.
	// Prerequisites:
	// Implementation must be concurrency safe; Non-blocking;
	// and must handle repetition of the same events (with some processing overhead).
	OnNewTcDiscovered(certificate *flow.TimeoutCertificate)
}

TimeoutCollectorConsumer consumes outbound notifications produced by HotStuff's timeout aggregation component. These events are primarily intended for the HotStuff-internal state machine (EventHandler), but might also be relevant to the larger node in which HotStuff is running.

Caution: the events are not strictly ordered by increasing views! The notifications are emitted by concurrent processing logic. Over larger time scales, the emitted events are for statistically increasing views. However, on short time scales there are _no_ monotonicity guarantees w.r.t. the events' views.

Implementations must:

• be concurrency safe
• be non-blocking
• handle repetition of the same events (with some processing overhead).

type TimeoutCollectorFactory

type TimeoutCollectorFactory interface {
	// Create is a factory method to generate a TimeoutCollector for a given view
	// Expected error returns during normal operations:
	//  * model.ErrViewForUnknownEpoch no epoch containing the given view is known
	// All other errors should be treated as exceptions.
	Create(view uint64) (TimeoutCollector, error)
}

TimeoutCollectorFactory performs creation of TimeoutCollector for a given view

type TimeoutCollectors

type TimeoutCollectors interface {
	// GetOrCreateCollector retrieves the TimeoutCollector for the specified
	// view or creates one if none exists.  When creating a timeout collector,
	// the view is used to query the consensus committee for the respective
	// Epoch the view belongs to.
	// It returns:
	//  -  (collector, true, nil) if no collector can be found by the view, and a new collector was created.
	//  -  (collector, false, nil) if the collector can be found by the view.
	//  -  (nil, false, error) if running into any exception creating the timeout collector.
	// Expected error returns during normal operations:
	//  * mempool.BelowPrunedThresholdError if view is below the pruning threshold
	//  * model.ErrViewForUnknownEpoch if view is not yet pruned but no epoch containing the given view is known
	GetOrCreateCollector(view uint64) (collector TimeoutCollector, created bool, err error)

	// PruneUpToView prunes the timeout collectors with views _below_ the given value, i.e.
	// we only retain and process timeout collectors, whose views are equal or larger than `lowestRetainedView`.
	// If `lowestRetainedView` is smaller than the previous value, the previous value is
	// kept and the method call is a NoOp.
	PruneUpToView(lowestRetainedView uint64)
}

TimeoutCollectors encapsulates the functionality to generate, store and prune `TimeoutCollector` instances (one per view). Its main purpose is to provide a higher-level API to `TimeoutAggregator` for managing and interacting with the view-specific `TimeoutCollector` instances. Implementations are concurrency safe.

type TimeoutProcessor

type TimeoutProcessor interface {
	// Process performs processing of single timeout object. This function is safe to call from multiple goroutines.
	// Expected error returns during normal operations:
	// * timeoutcollector.ErrTimeoutForIncompatibleView - submitted timeout for incompatible view
	// * model.InvalidTimeoutError - submitted invalid timeout(invalid structure or invalid signature)
	// * model.DuplicatedSignerError if a timeout from the same signer was previously already added
	//   It does _not necessarily_ imply that the timeout is invalid or the sender is equivocating.
	// All other errors should be treated as exceptions.
	Process(timeout *model.TimeoutObject) error
}

TimeoutProcessor ingests Timeout Objects [TO] for a particular view. It implements the algorithms for validating TOs, orchestrates their low-level aggregation and emits `OnPartialTcCreated` and `OnTcConstructedFromTimeouts` notifications. TimeoutProcessor cannot deduplicate TOs (this should be handled by the higher-level TimeoutCollector) and errors instead. Depending on their implementation, a TimeoutProcessor might drop timeouts or attempt to construct a TC.

type TimeoutProcessorFactory

type TimeoutProcessorFactory interface {
	// Create is a factory method to generate a TimeoutProcessor for a given view
	// Expected error returns during normal operations:
	//  * model.ErrViewForUnknownEpoch no epoch containing the given view is known
	// All other errors should be treated as exceptions.
	Create(view uint64) (TimeoutProcessor, error)
}

TimeoutProcessorFactory performs creation of TimeoutProcessor for a given view

type TimeoutSignatureAggregator

type TimeoutSignatureAggregator interface {
	// VerifyAndAdd verifies the signature under the stored public keys and adds the signature and the corresponding
	// highest QC to the internal set. Internal set and collected weight is modified iff signature _is_ valid.
	// The total weight of all collected signatures (excluding duplicates) is returned regardless
	// of any returned error.
	// Expected errors during normal operations:
	//  - model.InvalidSignerError if signerID is invalid (not a consensus participant)
	//  - model.DuplicatedSignerError if the signer has been already added
	//  - model.ErrInvalidSignature if signerID is valid but signature is cryptographically invalid
	VerifyAndAdd(signerID flow.Identifier, sig crypto.Signature, newestQCView uint64) (totalWeight uint64, exception error)

	// TotalWeight returns the total weight presented by the collected signatures.
	TotalWeight() uint64

	// View returns the view that this instance is aggregating signatures for.
	View() uint64

	// Aggregate aggregates the signatures and returns with additional data.
	// Aggregated signature will be returned as SigData of timeout certificate.
	// Caller can be sure that resulting signature is valid.
	// Expected errors during normal operations:
	//  - model.InsufficientSignaturesError if no signatures have been added yet
	Aggregate() (signersInfo []TimeoutSignerInfo, aggregatedSig crypto.Signature, exception error)
}

TimeoutSignatureAggregator aggregates timeout signatures for one particular view. When instantiating a TimeoutSignatureAggregator, the following information is supplied:

• The view for which the aggregator collects timeouts.
• For each replicas that is authorized to send a timeout at this particular view: the node ID, public staking keys, and weight

Timeouts for other views or from non-authorized replicas are rejected. In their TimeoutObjects, replicas include a signature over the pair (view, newestQCView), where `view` is the view number the timeout is for and `newestQCView` is the view of the newest QC known to the replica. TimeoutSignatureAggregator collects these signatures, internally tracks the total weight of all collected signatures. Note that in general the signed messages are different, which makes the aggregation a comparatively expensive operation. Upon calling `Aggregate`, the TimeoutSignatureAggregator aggregates all valid signatures collected up to this point. The aggregate signature is guaranteed to be correct, as only valid signatures are excepted as inputs. TimeoutSignatureAggregator internally tracks the total weight of all collected signatures. Implementations must be concurrency safe.

type TimeoutSignerInfo

type TimeoutSignerInfo struct {
	NewestQCView uint64
	Signer       flow.Identifier
}

TimeoutSignerInfo is a helper structure that stores the QC views that each signer contributed to a TC. Used as result of TimeoutSignatureAggregator.Aggregate()

type Validator

type Validator interface {

	// ValidateQC checks the validity of a QC.
	// During normal operations, the following error returns are expected:
	//  * model.InvalidQCError if the QC is invalid
	//  * model.ErrViewForUnknownEpoch if the QC refers unknown epoch
	ValidateQC(qc *flow.QuorumCertificate) error

	// ValidateTC checks the validity of a TC.
	// During normal operations, the following error returns are expected:
	//  * model.InvalidTCError if the TC is invalid
	//  * model.ErrViewForUnknownEpoch if the TC refers unknown epoch
	ValidateTC(tc *flow.TimeoutCertificate) error

	// ValidateProposal checks the validity of a proposal.
	// During normal operations, the following error returns are expected:
	//  * model.InvalidBlockError if the block is invalid
	//  * model.ErrViewForUnknownEpoch if the proposal refers unknown epoch
	ValidateProposal(proposal *model.Proposal) error

	// ValidateVote checks the validity of a vote.
	// Returns the full entity for the voter. During normal operations,
	// the following errors are expected:
	//  * model.InvalidVoteError for invalid votes
	//  * model.ErrViewForUnknownEpoch if the vote refers unknown epoch
	ValidateVote(vote *model.Vote) (*flow.Identity, error)
}

Validator provides functions to validate QC, proposals and votes.

type Verifier

type Verifier interface {

	// VerifyVote checks the cryptographic validity of a vote's `SigData` w.r.t.
	// the view and blockID. It is the responsibility of the calling code to ensure
	// that `voter` is authorized to vote.
	// Return values:
	//  * nil if `sigData` is cryptographically valid
	//  * model.InvalidFormatError if the signature has an incompatible format.
	//  * model.ErrInvalidSignature is the signature is invalid
	//  * model.InvalidSignerError is only relevant for extended signature schemes,
	//    where special signing authority is only given to a _subset_ of consensus
	//    participants (e.g. random beacon). In case a participant signed despite not
	//    being authorized, an InvalidSignerError is returned.
	//  * model.ErrViewForUnknownEpoch is only relevant for extended signature schemes,
	//    where querying of DKG might fail if no epoch containing the given view is known.
	//  * unexpected errors should be treated as symptoms of bugs or uncovered
	//    edge cases in the logic (i.e. as fatal)
	VerifyVote(voter *flow.Identity, sigData []byte, view uint64, blockID flow.Identifier) error

	// VerifyQC checks the cryptographic validity of a QC's `SigData` w.r.t. the
	// given view and blockID. It is the responsibility of the calling code to ensure that
	// all `signers` are authorized, without duplicates.
	// Return values:
	//  * nil if `sigData` is cryptographically valid
	//  * model.InvalidFormatError if `sigData` has an incompatible format
	//  * model.InsufficientSignaturesError if `signers is empty.
	//    Depending on the order of checks in the higher-level logic this error might
	//    be an indicator of a external byzantine input or an internal bug.
	//  * model.ErrInvalidSignature if a signature is invalid
	//  * model.InvalidSignerError is only relevant for extended signature schemes,
	//    where special signing authority is only given to a _subset_ of consensus
	//    participants (e.g. random beacon). In case a participant signed despite not
	//    being authorized, an InvalidSignerError is returned.
	//  * model.ErrViewForUnknownEpoch is only relevant for extended signature schemes,
	//    where querying of DKG might fail if no epoch containing the given view is known.
	//  * unexpected errors should be treated as symptoms of bugs or uncovered
	//	  edge cases in the logic (i.e. as fatal)
	VerifyQC(signers flow.IdentityList, sigData []byte, view uint64, blockID flow.Identifier) error

	// VerifyTC checks cryptographic validity of the TC's `sigData` w.r.t. the
	// given view. It is the responsibility of the calling code to ensure
	// that all `signers` are authorized, without duplicates. Return values:
	//  * nil if `sigData` is cryptographically valid
	//  * model.InsufficientSignaturesError if `signers is empty.
	//  * model.InvalidFormatError if `signers`/`highQCViews` have differing lengths
	//  * model.ErrInvalidSignature if a signature is invalid
	//  * unexpected errors should be treated as symptoms of bugs or uncovered
	//	  edge cases in the logic (i.e. as fatal)
	VerifyTC(signers flow.IdentityList, sigData []byte, view uint64, highQCViews []uint64) error
}

Verifier is the component responsible for the cryptographic integrity of votes, proposals and QC's against the block they are signing. Overall, there are two criteria for the validity of a vote and QC:

(1) the signer ID(s) must correspond to authorized consensus participants
(2) the signature must be cryptographically valid.

Note that Verifier only implements (2). This API design allows to decouple

 (i)  the common logic for checking that a super-majority of the consensus
      committee voted
 (ii) the handling of combined staking+RandomBeacon votes (consensus nodes)
	  vs only staking votes (collector nodes)

On the one hand, this API design makes code less concise, as the two checks are now distributed over API boundaries. On the other hand, we can avoid repeated Identity lookups in the implementation, which increases performance.

type VerifyingVoteProcessor

type VerifyingVoteProcessor interface {
	VoteProcessor

	// Block returns which block that will be used to collector votes for. Transition to VerifyingVoteCollector can occur only
	// when we have received block proposal so this information has to be available.
	Block() *model.Block
}

VerifyingVoteProcessor is a VoteProcessor that attempts to construct a QC for the given block.

type VoteAggregator

type VoteAggregator interface {
	module.ReadyDoneAware
	module.Startable

	// AddVote verifies and aggregates a vote.
	// The voting block could either be known or unknown.
	// If the voting block is unknown, the vote won't be processed until AddBlock is called with the block.
	// This method can be called concurrently, votes will be queued and processed asynchronously.
	AddVote(vote *model.Vote)

	// AddBlock notifies the VoteAggregator that it should start processing votes for the given block.
	// The input block is queued internally within the `VoteAggregator` and processed _asynchronously_
	// by the VoteAggregator's internal worker routines.
	// CAUTION: we expect that the input block's validity has been confirmed prior to calling AddBlock,
	// including the proposer's signature. Otherwise, VoteAggregator might crash or exhibit undefined
	// behaviour.
	AddBlock(block *model.Proposal)

	// InvalidBlock notifies the VoteAggregator about an invalid proposal, so that it
	// can process votes for the invalid block and slash the voters.
	// No errors are expected during normal operations
	InvalidBlock(block *model.Proposal) error

	// PruneUpToView deletes all votes _below_ to the given view, as well as
	// related indices. We only retain and process whose view is equal or larger
	// than `lowestRetainedView`. If `lowestRetainedView` is smaller than the
	// previous value, the previous value is kept and the method call is a NoOp.
	PruneUpToView(view uint64)
}

VoteAggregator verifies and aggregates votes to build QC. When enough votes have been collected, it builds a QC and send it to the EventLoop VoteAggregator also detects protocol violation, including invalid votes, double voting etc, and notifies a HotStuff consumer for slashing.

type VoteCollector

type VoteCollector interface {
	// ProcessBlock performs validation of block signature and processes block with respected collector.
	// Calling this function will mark conflicting collector as stale and change state of valid collectors
	// It returns nil if the block is valid.
	// It returns model.InvalidBlockError if block is invalid.
	// It returns other error if there is exception processing the block.
	ProcessBlock(block *model.Proposal) error

	// AddVote adds a vote to the collector
	// When enough votes have been added to produce a QC, the QC will be created asynchronously, and
	// passed to EventLoop through a callback.
	// No errors are expected during normal operations.
	AddVote(vote *model.Vote) error

	// RegisterVoteConsumer registers a VoteConsumer. Upon registration, the collector
	// feeds all cached votes into the consumer in the order they arrived.
	// CAUTION, VoteConsumer implementations must be
	//  * NON-BLOCKING and consume the votes without noteworthy delay, and
	//  * CONCURRENCY SAFE
	RegisterVoteConsumer(consumer VoteConsumer)

	// View returns the view that this instance is collecting votes for.
	// This method is useful when adding the newly created vote collector to vote collectors map.
	View() uint64

	// Status returns the status of the vote collector
	Status() VoteCollectorStatus
}

VoteCollector collects all votes for a specified view. On the happy path, it generates a QC when enough votes have been collected. The VoteCollector internally delegates the vote-format specific processing to the VoteProcessor.

type VoteCollectorStatus

type VoteCollectorStatus int

VoteCollectorStatus indicates the VoteCollector's status It has three different status.

const (
	// VoteCollectorStatusCaching is for the status when the block has not been received.
	// The vote collector in this status will cache all the votes without verifying them
	VoteCollectorStatusCaching VoteCollectorStatus = iota

	// VoteCollectorStatusVerifying is for the status when the block has been received,
	// and is able to process all votes for it.
	VoteCollectorStatusVerifying

	// VoteCollectorStatusInvalid is for the status when the block has been verified and
	// is invalid. All votes to this block will be collected to slash the voter.
	VoteCollectorStatusInvalid
)

func (VoteCollectorStatus) String

func (ps VoteCollectorStatus) String() string

type VoteCollectors

type VoteCollectors interface {
	module.ReadyDoneAware
	module.Startable

	// GetOrCreateCollector retrieves the hotstuff.VoteCollector for the specified
	// view or creates one if none exists.
	// When creating a vote collector, the view will be used to get epoch by view, then create the random beacon
	// signer object by epoch, because epoch determines DKG, which determines random beacon committee.
	// It returns:
	//  -  (collector, true, nil) if no collector can be found by the view, and a new collector was created.
	//  -  (collector, false, nil) if the collector can be found by the view
	//  -  (nil, false, error) if running into any exception creating the vote collector state machine
	// Expected error returns during normal operations:
	//  * mempool.BelowPrunedThresholdError - in case view is lower than last pruned view
	GetOrCreateCollector(view uint64) (collector VoteCollector, created bool, err error)

	// PruneUpToView prunes the vote collectors with views _below_ the given value, i.e.
	// we only retain and process whose view is equal or larger than `lowestRetainedView`.
	// If `lowestRetainedView` is smaller than the previous value, the previous value is
	// kept and the method call is a NoOp.
	PruneUpToView(lowestRetainedView uint64)
}

VoteCollectors is an interface which allows VoteAggregator to interact with collectors structured by view. Implementations of this interface are responsible for state transitions of `VoteCollector`s and pruning of stale and outdated collectors by view.

type VoteConsumer

type VoteConsumer func(vote *model.Vote)

VoteConsumer consumes all votes for one specific view. It is registered with the `VoteCollector` for the respective view. Upon registration, the `VoteCollector` feeds votes into the consumer in the order they are received (already cached votes as well as votes received in the future). Only votes that pass de-duplication and equivocation detection are passed on. CAUTION, VoteConsumer implementations must be

• NON-BLOCKING and consume the votes without noteworthy delay, and
• CONCURRENCY SAFE

type VoteProcessor

type VoteProcessor interface {
	// Process performs processing of single vote. This function is safe to call from multiple goroutines.
	// Expected error returns during normal operations:
	// * VoteForIncompatibleBlockError - submitted vote for incompatible block
	// * VoteForIncompatibleViewError - submitted vote for incompatible view
	// * model.InvalidVoteError - submitted vote with invalid signature
	// * model.DuplicatedSignerError - vote from a signer whose vote was previously already processed
	// All other errors should be treated as exceptions.
	Process(vote *model.Vote) error

	// Status returns the status of the vote processor
	Status() VoteCollectorStatus
}

VoteProcessor processes votes. It implements the vote-format specific processing logic. Depending on their implementation, a VoteProcessor might drop votes or attempt to construct a QC.

type VoteProcessorFactory

type VoteProcessorFactory interface {
	// Create instantiates a VerifyingVoteProcessor for processing votes for a specific proposal.
	// Caller can be sure that proposal vote was successfully verified and processed.
	// Expected error returns during normal operations:
	// * model.InvalidBlockError - proposal has invalid proposer vote
	Create(log zerolog.Logger, proposal *model.Proposal) (VerifyingVoteProcessor, error)
}

VoteProcessorFactory is a factory that can be used to create a verifying vote processors for a specific proposal. Depending on factory implementation it will return processors for consensus or collection clusters

type WeightedSignatureAggregator

type WeightedSignatureAggregator interface {
	// Verify verifies the signature under the stored public keys and message.
	// Expected errors during normal operations:
	//  - model.InvalidSignerError if signerID is invalid (not a consensus participant)
	//  - model.ErrInvalidSignature if signerID is valid but signature is cryptographically invalid
	Verify(signerID flow.Identifier, sig crypto.Signature) error

	// TrustedAdd adds a signature to the internal set of signatures and adds the signer's
	// weight to the total collected weight, iff the signature is _not_ a duplicate. The
	// total weight of all collected signatures (excluding duplicates) is returned regardless
	// of any returned error.
	// Expected errors during normal operations:
	//  - model.InvalidSignerError if signerID is invalid (not a consensus participant)
	//  - model.DuplicatedSignerError if the signer has been already added
	TrustedAdd(signerID flow.Identifier, sig crypto.Signature) (totalWeight uint64, exception error)

	// TotalWeight returns the total weight presented by the collected signatures.
	TotalWeight() uint64

	// Aggregate aggregates the signatures and returns the aggregated signature.
	// The function performs a final verification and errors if the aggregated
	// signature is not valid. This is required for the function safety since
	// `TrustedAdd` allows adding invalid signatures.
	// Expected errors during normal operations:
	//  - model.InsufficientSignaturesError if no signatures have been added yet
	//  - model.InvalidSignatureIncludedError if some signature(s), included via TrustedAdd, are invalid
	Aggregate() (flow.IdentifierList, []byte, error)
}

WeightedSignatureAggregator aggregates signatures of the same signature scheme and the same message from different signers. The public keys and message are agreed upon upfront. It is also recommended to only aggregate signatures generated with keys representing equivalent security-bit level. Furthermore, a weight [unsigned int64] is assigned to each signer ID. The WeightedSignatureAggregator internally tracks the total weight of all collected signatures. Implementations must be concurrency safe.

type Workerpool

type Workerpool interface {
	Workers

	// StopWait stops the worker pool and waits for all queued tasks to
	// complete.  No additional tasks may be submitted, but all pending tasks are
	// executed by workers before this function returns.
	StopWait()
}

Workerpool adds the functionality to terminate the workers to the Workers interface.

type Workers

type Workers interface {
	// Submit enqueues a function for a worker to execute. Submit will not block
	// regardless of the number of tasks submitted. Each task is immediately
	// given to an available worker or queued otherwise. Tasks are processed in
	// FiFO order.
	Submit(task func())
}

Workers queues and processes submitted tasks. We explicitly do not expose any functionality to terminate the worker pool.