raft

type AppendEntriesRequest struct {
	RPCHeader

	// Provide the current term and leader
	Term uint64

	// Deprecated: use RPCHeader.Addr instead
	Leader []byte

	// Provide the previous entries for integrity checking
	PrevLogEntry uint64
	PrevLogTerm  uint64

	// New entries to commit
	Entries []*Log

	// Commit index on the leader
	LeaderCommitIndex uint64
}

type AppendEntriesResponse struct {
	RPCHeader

	// Newer term if leader is out of date
	Term uint64

	// Last Log is a hint to help accelerate rebuilding slow nodes
	LastLog uint64

	// We may not succeed if we have a conflicting entry
	Success bool

	// There are scenarios where this request didn't succeed
	// but there's no need to wait/back-off the next attempt.
	NoRetryBackoff bool
}

type AppendFuture interface {
	Future

	// Start returns the time that the append request was started.
	// It is always OK to call this method.
	Start() time.Time

	// Request holds the parameters of the AppendEntries call.
	// It is always OK to call this method.
	Request() *AppendEntriesRequest

	// Response holds the results of the AppendEntries call.
	// This method must only be called after the Error
	// method returns, and will only be valid on success.
	Response() *AppendEntriesResponse
}

type AppendPipeline interface {
	// AppendEntries is used to add another request to the pipeline.
	// The send may block which is an effective form of back-pressure.
	AppendEntries(args *AppendEntriesRequest, resp *AppendEntriesResponse) (AppendFuture, error)

	// Consumer returns a channel that can be used to consume
	// response futures when they are ready.
	Consumer() <-chan AppendFuture

	// Close closes the pipeline and cancels all inflight RPCs
	Close() error
}

type ApplyFuture interface {
	IndexFuture

	// Response returns the FSM response as returned by the FSM.Apply method. This
	// must not be called until after the Error method has returned.
	// Note that if FSM.Apply returns an error, it will be returned by Response,
	// and not by the Error method, so it is always important to check Response
	// for errors from the FSM.
	Response() interface{}
}

type BatchingFSM interface {
	// ApplyBatch is invoked once a batch of log entries has been committed and
	// are ready to be applied to the FSM. ApplyBatch will take in an array of
	// log entries. These log entries will be in the order they were committed,
	// will not have gaps, and could be of a few log types. Clients should check
	// the log type prior to attempting to decode the data attached. Presently
	// the LogCommand and LogConfiguration types will be sent.
	//
	// The returned slice must be the same length as the input and each response
	// should correlate to the log at the same index of the input. The returned
	// values will be made available in the ApplyFuture returned by Raft.Apply
	// method if that method was called on the same Raft node as the FSM.
	ApplyBatch([]*Log) []interface{}

	FSM
}

type Config struct {
	// ProtocolVersion allows a Raft server to inter-operate with older
	// Raft servers running an older version of the code. This is used to
	// version the wire protocol as well as Raft-specific log entries that
	// the server uses when _speaking_ to other servers. There is currently
	// no auto-negotiation of versions so all servers must be manually
	// configured with compatible versions. See ProtocolVersionMin and
	// ProtocolVersionMax for the versions of the protocol that this server
	// can _understand_.
	ProtocolVersion ProtocolVersion

	// HeartbeatTimeout specifies the time in follower state without contact
	// from a leader before we attempt an election.
	HeartbeatTimeout time.Duration

	// ElectionTimeout specifies the time in candidate state without contact
	// from a leader before we attempt an election.
	ElectionTimeout time.Duration

	// CommitTimeout specifies the time without an Apply operation before the
	// leader sends an AppendEntry RPC to followers, to ensure a timely commit of
	// log entries.
	// Due to random staggering, may be delayed as much as 2x this value.
	CommitTimeout time.Duration

	// MaxAppendEntries controls the maximum number of append entries
	// to send at once. We want to strike a balance between efficiency
	// and avoiding waste if the follower is going to reject because of
	// an inconsistent log.
	MaxAppendEntries int

	// BatchApplyCh indicates whether we should buffer applyCh
	// to size MaxAppendEntries. This enables batch log commitment,
	// but breaks the timeout guarantee on Apply. Specifically,
	// a log can be added to the applyCh buffer but not actually be
	// processed until after the specified timeout.
	BatchApplyCh bool

	// If we are a member of a cluster, and RemovePeer is invoked for the
	// local node, then we forget all peers and transition into the follower state.
	// If ShutdownOnRemove is set, we additional shutdown Raft. Otherwise,
	// we can become a leader of a cluster containing only this node.
	ShutdownOnRemove bool

	// TrailingLogs controls how many logs we leave after a snapshot. This is used
	// so that we can quickly replay logs on a follower instead of being forced to
	// send an entire snapshot. The value passed here is the initial setting used.
	// This can be tuned during operation using ReloadConfig.
	TrailingLogs uint64

	// SnapshotInterval controls how often we check if we should perform a
	// snapshot. We randomly stagger between this value and 2x this value to avoid
	// the entire cluster from performing a snapshot at once. The value passed
	// here is the initial setting used. This can be tuned during operation using
	// ReloadConfig.
	SnapshotInterval time.Duration

	// SnapshotThreshold controls how many outstanding logs there must be before
	// we perform a snapshot. This is to prevent excessive snapshotting by
	// replaying a small set of logs instead. The value passed here is the initial
	// setting used. This can be tuned during operation using ReloadConfig.
	SnapshotThreshold uint64

	// LeaderLeaseTimeout is used to control how long the "lease" lasts
	// for being the leader without being able to contact a quorum
	// of nodes. If we reach this interval without contact, we will
	// step down as leader.
	LeaderLeaseTimeout time.Duration

	// LocalID is a unique ID for this server across all time. When running with
	// ProtocolVersion < 3, you must set this to be the same as the network
	// address of your transport.
	LocalID ServerID

	// NotifyCh is used to provide a channel that will be notified of leadership
	// changes. Raft will block writing to this channel, so it should either be
	// buffered or aggressively consumed.
	NotifyCh chan<- bool

	// LogOutput is used as a sink for logs, unless Logger is specified.
	// Defaults to os.Stderr.
	LogOutput io.Writer

	// LogLevel represents a log level. If the value does not match a known
	// logging level hclog.NoLevel is used.
	LogLevel string

	// Logger is a user-provided logger. If nil, a logger writing to
	// LogOutput with LogLevel is used.
	Logger hclog.Logger

	// NoSnapshotRestoreOnStart controls if raft will restore a snapshot to the
	// FSM on start. This is useful if your FSM recovers from other mechanisms
	// than raft snapshotting. Snapshot metadata will still be used to initialize
	// raft's configuration and index values.
	NoSnapshotRestoreOnStart bool
	// contains filtered or unexported fields
}

type Configuration struct {
	Servers []Server
}

type ConfigurationChangeCommand uint8

type ConfigurationFuture interface {
	IndexFuture

	// Configuration contains the latest configuration. This must
	// not be called until after the Error method has returned.
	Configuration() Configuration
}

type ConfigurationStore interface {
	// ConfigurationStore is a superset of the FSM functionality
	FSM

	// StoreConfiguration is invoked once a log entry containing a configuration
	// change is committed. It takes the index at which the configuration was
	// written and the configuration value.
	StoreConfiguration(index uint64, configuration Configuration)
}

type CountingReader interface {
	io.Reader
	Count() int64
}

type DiscardSnapshotSink struct{}

type DiscardSnapshotStore struct{}

type FSM interface {
	// Apply is called once a log entry is committed by a majority of the cluster.
	//
	// Apply should apply the log to the FSM. Apply must be deterministic and
	// produce the same result on all peers in the cluster.
	//
	// The returned value is returned to the client as the ApplyFuture.Response.
	Apply(*Log) interface{}

	// Snapshot returns an FSMSnapshot used to: support log compaction, to
	// restore the FSM to a previous state, or to bring out-of-date followers up
	// to a recent log index.
	//
	// The Snapshot implementation should return quickly, because Apply can not
	// be called while Snapshot is running. Generally this means Snapshot should
	// only capture a pointer to the state, and any expensive IO should happen
	// as part of FSMSnapshot.Persist.
	//
	// Apply and Snapshot are always called from the same thread, but Apply will
	// be called concurrently with FSMSnapshot.Persist. This means the FSM should
	// be implemented to allow for concurrent updates while a snapshot is happening.
	Snapshot() (FSMSnapshot, error)

	// Restore is used to restore an FSM from a snapshot. It is not called
	// concurrently with any other command. The FSM must discard all previous
	// state before restoring the snapshot.
	Restore(snapshot io.ReadCloser) error
}

type FSMSnapshot interface {
	// Persist should dump all necessary state to the WriteCloser 'sink',
	// and call sink.Close() when finished or call sink.Cancel() on error.
	Persist(sink SnapshotSink) error

	// Release is invoked when we are finished with the snapshot.
	Release()
}

type FailedHeartbeatObservation struct {
	PeerID      ServerID
	LastContact time.Time
}

type FileSnapshotSink struct {
	// contains filtered or unexported fields
}

type FileSnapshotStore struct {
	// contains filtered or unexported fields
}

type FilterFn func(o *Observation) bool

type Future interface {
	// Error blocks until the future arrives and then returns the error status
	// of the future. This may be called any number of times - all calls will
	// return the same value, however is not OK to call this method twice
	// concurrently on the same Future instance.
	// Error will only return generic errors related to raft, such
	// as ErrLeadershipLost, or ErrRaftShutdown. Some operations, such as
	// ApplyLog, may also return errors from other methods.
	Error() error
}

type IndexFuture interface {
	Future

	// Index holds the index of the newly applied log entry.
	// This must not be called until after the Error method has returned.
	Index() uint64
}

type InmemSnapshotSink struct {
	// contains filtered or unexported fields
}

type InmemSnapshotStore struct {
	sync.RWMutex
	// contains filtered or unexported fields
}

type InmemStore struct {
	// contains filtered or unexported fields
}

type InmemTransport struct {
	sync.RWMutex
	// contains filtered or unexported fields
}

type InstallSnapshotRequest struct {
	RPCHeader
	SnapshotVersion SnapshotVersion

	Term   uint64
	Leader []byte

	// These are the last index/term included in the snapshot
	LastLogIndex uint64
	LastLogTerm  uint64

	// Peer Set in the snapshot.
	// but remains here in case we receive an InstallSnapshot from a leader
	// that's running old code.
	// Deprecated: This is deprecated in favor of Configuration
	Peers []byte

	// Cluster membership.
	Configuration []byte
	// Log index where 'Configuration' entry was originally written.
	ConfigurationIndex uint64

	// Size of the snapshot
	Size int64
}

type InstallSnapshotResponse struct {
	RPCHeader

	Term    uint64
	Success bool
}

type LeaderObservation struct {
	// DEPRECATED The LeaderAddr field should now be used
	Leader     ServerAddress
	LeaderAddr ServerAddress
	LeaderID   ServerID
}

type LeadershipTransferFuture interface {
	Future
}

type Log struct {
	// Index holds the index of the log entry.
	Index uint64

	// Term holds the election term of the log entry.
	Term uint64

	// Type holds the type of the log entry.
	Type LogType

	// Data holds the log entry's type-specific data.
	Data []byte

	// Extensions holds an opaque byte slice of information for middleware. It
	// is up to the client of the library to properly modify this as it adds
	// layers and remove those layers when appropriate. This value is a part of
	// the log, so very large values could cause timing issues.
	//
	// N.B. It is _up to the client_ to handle upgrade paths. For instance if
	// using this with go-raftchunking, the client should ensure that all Raft
	// peers are using a version that can handle that extension before ever
	// actually triggering chunking behavior. It is sometimes sufficient to
	// ensure that non-leaders are upgraded first, then the current leader is
	// upgraded, but a leader changeover during this process could lead to
	// trouble, so gating extension behavior via some flag in the client
	// program is also a good idea.
	Extensions []byte

	// AppendedAt stores the time the leader first appended this log to it's
	// LogStore. Followers will observe the leader's time. It is not used for
	// coordination or as part of the replication protocol at all. It exists only
	// to provide operational information for example how many seconds worth of
	// logs are present on the leader which might impact follower's ability to
	// catch up after restoring a large snapshot. We should never rely on this
	// being in the past when appending on a follower or reading a log back since
	// the clock skew can mean a follower could see a log with a future timestamp.
	// In general too the leader is not required to persist the log before
	// delivering to followers although the current implementation happens to do
	// this.
	AppendedAt time.Time
}

type LogCache struct {
	// contains filtered or unexported fields
}

type LogStore interface {
	// FirstIndex returns the first index written. 0 for no entries.
	FirstIndex() (uint64, error)

	// LastIndex returns the last index written. 0 for no entries.
	LastIndex() (uint64, error)

	// GetLog gets a log entry at a given index.
	GetLog(index uint64, log *Log) error

	// StoreLog stores a log entry.
	StoreLog(log *Log) error

	// StoreLogs stores multiple log entries. By default the logs stored may not be contiguous with previous logs (i.e. may have a gap in Index since the last log written). If an implementation can't tolerate this it may optionally implement `MonotonicLogStore` to indicate that this is not allowed. This changes Raft's behaviour after restoring a user snapshot to remove all previous logs instead of relying on a "gap" to signal the discontinuity between logs before the snapshot and logs after.
	StoreLogs(logs []*Log) error

	// DeleteRange deletes a range of log entries. The range is inclusive.
	DeleteRange(min, max uint64) error
}

type LogType uint8

type LoopbackTransport interface {
	Transport // Embedded transport reference
	WithPeers // Embedded peer management
	WithClose // with a close routine
}

type MakeClusterOpts struct {
	Peers           int
	Bootstrap       bool
	Conf            *Config
	ConfigStoreFSM  bool
	MakeFSMFunc     func() FSM
	LongstopTimeout time.Duration
	MonotonicLogs   bool
}

type MockFSM struct {
	sync.Mutex
	// contains filtered or unexported fields
}

type MockFSMConfigStore struct {
	FSM
}

type MockMonotonicLogStore struct {
	// contains filtered or unexported fields
}

type MockSnapshot struct {
	// contains filtered or unexported fields
}

type MonotonicLogStore interface {
	IsMonotonic() bool
}

type NetworkTransport struct {
	TimeoutScale int
	// contains filtered or unexported fields
}

type NetworkTransportConfig struct {
	// ServerAddressProvider is used to override the target address when establishing a connection to invoke an RPC
	ServerAddressProvider ServerAddressProvider

	Logger hclog.Logger

	// Dialer
	Stream StreamLayer

	// MaxPool controls how many connections we will pool
	MaxPool int

	// MaxRPCsInFlight controls the pipelining "optimization" when replicating
	// entries to followers.
	//
	// Setting this to 1 explicitly disables pipelining since no overlapping of
	// request processing is allowed. If set to 1 the pipelining code path is
	// skipped entirely and every request is entirely synchronous.
	//
	// If zero is set (or left as default), DefaultMaxRPCsInFlight is used which
	// is currently 2. A value of 2 overlaps the preparation and sending of the
	// next request while waiting for the previous response, but avoids additional
	// queuing.
	//
	// Historically this was internally fixed at (effectively) 130 however
	// performance testing has shown that in practice the pipelining optimization
	// combines badly with batching and actually has a very large negative impact
	// on commit latency when throughput is high, whilst having very little
	// benefit on latency or throughput in any other case! See
	// [#541](https://github.com/hashicorp/raft/pull/541) for more analysis of the
	// performance impacts.
	//
	// Increasing this beyond 2 is likely to be beneficial only in very
	// high-latency network conditions. HashiCorp doesn't recommend using our own
	// products this way.
	//
	// To maintain the behavior from before version 1.4.1 exactly, set this to
	// 130. The old internal constant was 128 but was used directly as a channel
	// buffer size. Since we send before blocking on the channel and unblock the
	// channel as soon as the receiver is done with the earliest outstanding
	// request, even an unbuffered channel (buffer=0) allows one request to be
	// sent while waiting for the previous one (i.e. 2 inflight). so the old
	// buffer actually allowed 130 RPCs to be inflight at once.
	MaxRPCsInFlight int

	// Timeout is used to apply I/O deadlines. For InstallSnapshot, we multiply
	// the timeout by (SnapshotSize / TimeoutScale).
	Timeout time.Duration

	// MsgpackUseNewTimeFormat when set to true, force the underlying msgpack
	// codec to use the new format of time.Time when encoding (used in
	// go-msgpack v1.1.5 by default). Decoding is not affected, as all
	// go-msgpack v2.1.0+ decoders know how to decode both formats.
	MsgpackUseNewTimeFormat bool
}

type Observation struct {
	// Raft holds the Raft instance generating the observation.
	Raft *Raft
	// Data holds observation-specific data. Possible types are
	// RequestVoteRequest
	// RaftState
	// PeerObservation
	// LeaderObservation
	Data interface{}
}

type Observer struct {
	// contains filtered or unexported fields
}

type PeerObservation struct {
	Removed bool
	Peer    Server
}

type ProtocolVersion int

type RPC struct {
	Command  interface{}
	Reader   io.Reader // Set only for InstallSnapshot
	RespChan chan<- RPCResponse
}

type RPCHeader struct {
	// ProtocolVersion is the version of the protocol the sender is
	// speaking.
	ProtocolVersion ProtocolVersion
	// ID is the ServerID of the node sending the RPC Request or Response
	ID []byte
	// Addr is the ServerAddr of the node sending the RPC Request or Response
	Addr []byte
}

type RPCResponse struct {
	Response interface{}
	Error    error
}

type Raft struct {
	// contains filtered or unexported fields
}

type RaftState uint32

type ReadCloserWrapper interface {
	io.ReadCloser
	WrappedReadCloser() io.ReadCloser
}

type ReloadableConfig struct {
	// TrailingLogs controls how many logs we leave after a snapshot. This is used
	// so that we can quickly replay logs on a follower instead of being forced to
	// send an entire snapshot. The value passed here updates the setting at runtime
	// which will take effect as soon as the next snapshot completes and truncation
	// occurs.
	TrailingLogs uint64

	// SnapshotInterval controls how often we check if we should perform a snapshot.
	// We randomly stagger between this value and 2x this value to avoid the entire
	// cluster from performing a snapshot at once.
	SnapshotInterval time.Duration

	// SnapshotThreshold controls how many outstanding logs there must be before
	// we perform a snapshot. This is to prevent excessive snapshots when we can
	// just replay a small set of logs.
	SnapshotThreshold uint64

	// HeartbeatTimeout specifies the time in follower state without
	// a leader before we attempt an election.
	HeartbeatTimeout time.Duration

	// ElectionTimeout specifies the time in candidate state without
	// a leader before we attempt an election.
	ElectionTimeout time.Duration
}

type RequestVoteRequest struct {
	RPCHeader

	// Provide the term and our id
	Term uint64

	// Deprecated: use RPCHeader.Addr instead
	Candidate []byte

	// Used to ensure safety
	LastLogIndex uint64
	LastLogTerm  uint64

	// Used to indicate to peers if this vote was triggered by a leadership
	// transfer. It is required for leadership transfer to work, because servers
	// wouldn't vote otherwise if they are aware of an existing leader.
	LeadershipTransfer bool
}

type RequestVoteResponse struct {
	RPCHeader

	// Newer term if leader is out of date.
	Term uint64

	// Peers is deprecated, but required by servers that only understand
	// protocol version 0. This is not populated in protocol version 2
	// and later.
	Peers []byte

	// Is the vote granted.
	Granted bool
}

type ResumedHeartbeatObservation struct {
	PeerID ServerID
}

type Server struct {
	// Suffrage determines whether the server gets a vote.
	Suffrage ServerSuffrage
	// ID is a unique string identifying this server for all time.
	ID ServerID
	// Address is its network address that a transport can contact.
	Address ServerAddress
}

type ServerAddress string

type ServerAddressProvider interface {
	ServerAddr(id ServerID) (ServerAddress, error)
}

type ServerID string

type ServerSuffrage int

type SnapshotFuture interface {
	Future

	// Open is a function you can call to access the underlying snapshot and
	// its metadata. This must not be called until after the Error method
	// has returned.
	Open() (*SnapshotMeta, io.ReadCloser, error)
}

type SnapshotMeta struct {
	// Version is the version number of the snapshot metadata. This does not cover
	// the application's data in the snapshot, that should be versioned
	// separately.
	Version SnapshotVersion

	// ID is opaque to the store, and is used for opening.
	ID string

	// Index and Term store when the snapshot was taken.
	Index uint64
	Term  uint64

	// Peers is deprecated and used to support version 0 snapshots, but will
	// be populated in version 1 snapshots as well to help with upgrades.
	Peers []byte

	// Configuration and ConfigurationIndex are present in version 1
	// snapshots and later.
	Configuration      Configuration
	ConfigurationIndex uint64

	// Size is the size of the snapshot in bytes.
	Size int64
}

type SnapshotSink interface {
	io.WriteCloser
	ID() string
	Cancel() error
}

type SnapshotStore interface {
	// Create is used to begin a snapshot at a given index and term, and with
	// the given committed configuration. The version parameter controls
	// which snapshot version to create.
	Create(version SnapshotVersion, index, term uint64, configuration Configuration,
		configurationIndex uint64, trans Transport) (SnapshotSink, error)

	// List is used to list the available snapshots in the store.
	// It should return then in descending order, with the highest index first.
	List() ([]*SnapshotMeta, error)

	// Open takes a snapshot ID and provides a ReadCloser. Once close is
	// called it is assumed the snapshot is no longer needed.
	Open(id string) (*SnapshotMeta, io.ReadCloser, error)
}

type SnapshotVersion int

type StableStore interface {
	Set(key []byte, val []byte) error

	// Get returns the value for key, or an empty byte slice if key was not found.
	Get(key []byte) ([]byte, error)

	SetUint64(key []byte, val uint64) error

	// GetUint64 returns the uint64 value for key, or 0 if key was not found.
	GetUint64(key []byte) (uint64, error)
}

type StreamLayer interface {
	net.Listener

	// Dial is used to create a new outgoing connection
	Dial(address ServerAddress, timeout time.Duration) (net.Conn, error)
}

type TCPStreamLayer struct {
	// contains filtered or unexported fields
}

type TimeoutNowRequest struct {
	RPCHeader
}

type TimeoutNowResponse struct {
	RPCHeader
}

type Transport interface {
	// Consumer returns a channel that can be used to
	// consume and respond to RPC requests.
	Consumer() <-chan RPC

	// LocalAddr is used to return our local address to distinguish from our peers.
	LocalAddr() ServerAddress

	// AppendEntriesPipeline returns an interface that can be used to pipeline
	// AppendEntries requests.
	AppendEntriesPipeline(id ServerID, target ServerAddress) (AppendPipeline, error)

	// AppendEntries sends the appropriate RPC to the target node.
	AppendEntries(id ServerID, target ServerAddress, args *AppendEntriesRequest, resp *AppendEntriesResponse) error

	// RequestVote sends the appropriate RPC to the target node.
	RequestVote(id ServerID, target ServerAddress, args *RequestVoteRequest, resp *RequestVoteResponse) error

	// InstallSnapshot is used to push a snapshot down to a follower. The data is read from
	// the ReadCloser and streamed to the client.
	InstallSnapshot(id ServerID, target ServerAddress, args *InstallSnapshotRequest, resp *InstallSnapshotResponse, data io.Reader) error

	// EncodePeer is used to serialize a peer's address.
	EncodePeer(id ServerID, addr ServerAddress) []byte

	// DecodePeer is used to deserialize a peer's address.
	DecodePeer([]byte) ServerAddress

	// SetHeartbeatHandler is used to setup a heartbeat handler
	// as a fast-pass. This is to avoid head-of-line blocking from
	// disk IO. If a Transport does not support this, it can simply
	// ignore the call, and push the heartbeat onto the Consumer channel.
	SetHeartbeatHandler(cb func(rpc RPC))

	// TimeoutNow is used to start a leadership transfer to the target node.
	TimeoutNow(id ServerID, target ServerAddress, args *TimeoutNowRequest, resp *TimeoutNowResponse) error
}

type WithClose interface {
	// Close permanently closes a transport, stopping
	// any associated goroutines and freeing other resources.
	Close() error
}

type WithPeers interface {
	Connect(peer ServerAddress, t Transport) // Connect a peer
	Disconnect(peer ServerAddress)           // Disconnect a given peer
	DisconnectAll()                          // Disconnect all peers, possibly to reconnect them later
}

type WithRPCHeader interface {
	GetRPCHeader() RPCHeader
}

type WrappingFSM interface {
	Underlying() FSM
}