早在2013年11月份,在raft論文還只能在網上下載到草稿版時,我曾經寫過一篇blog對其進行簡要分析。4年過去了,各種raft協議的講解鋪天蓋地,raft也確實得到了廣泛的應用。其中最知名的應用莫過於etcd。etcd將raft協議本身實現為一個library,位於https://github.com/coreos/etcd/tree/master/raft,然后本身作為一個應用使用它。
本文不講解raft協議核心內容,而是站在一個etcd raft library使用者的角度,講解要用上這個library需要了解的東西。
這個library使用起來相對來說還是有點麻煩。官方有一個使用示例在 https://github.com/coreos/etcd/tree/master/contrib/raftexample。整體來說,這個庫實現了raft協議核心的內容,比如append log的邏輯,選主邏輯,snapshot,成員變更等邏輯。需要明確的是:library沒有實現消息的網絡傳輸和接收,庫只會把一些待發送的消息保存在內存中,用戶自定義的網絡傳輸層取出消息並發送出去,並且在網絡接收端,需要調一個library的函數,用於將收到的消息傳入library,后面會詳細說明。同時,library定義了一個Storage接口,需要library的使用者自行實現。
Storage接口如下:
// Storage is an interface that may be implemented by the application
// to retrieve log entries from storage.
//
// If any Storage method returns an error, the raft instance will
// become inoperable and refuse to participate in elections; the
// application is responsible for cleanup and recovery in this case.
type Storage interface {
// InitialState returns the saved HardState and ConfState information.
InitialState() (pb.HardState, pb.ConfState, error)
// Entries returns a slice of log entries in the range [lo,hi).
// MaxSize limits the total size of the log entries returned, but
// Entries returns at least one entry if any.
Entries(lo, hi, maxSize uint64) ([]pb.Entry, error)
// Term returns the term of entry i, which must be in the range
// [FirstIndex()-1, LastIndex()]. The term of the entry before
// FirstIndex is retained for matching purposes even though the
// rest of that entry may not be available.
Term(i uint64) (uint64, error)
// LastIndex returns the index of the last entry in the log.
LastIndex() (uint64, error)
// FirstIndex returns the index of the first log entry that is
// possibly available via Entries (older entries have been incorporated
// into the latest Snapshot; if storage only contains the dummy entry the
// first log entry is not available).
FirstIndex() (uint64, error)
// Snapshot returns the most recent snapshot.
// If snapshot is temporarily unavailable, it should return ErrSnapshotTemporarilyUnavailable,
// so raft state machine could know that Storage needs some time to prepare
// snapshot and call Snapshot later.
Snapshot() (pb.Snapshot, error)
}
這些接口在library中會被用到。熟悉raft協議的人不難理解。上面提到的官方示例https://github.com/coreos/etcd/tree/master/contrib/raftexample中使用了library自帶的MemoryStorage,和etcd的wal和snap包做持久化,重啟的時候從wal和snap中獲取日志恢復MemoryStorage。
要提供這種IO/網絡密集型的東西,提高吞吐最好的手段就是batch加批處理了。etcd raft library正是這么做的。
下面看一下為了做這事,etcd提供的核心抽象Ready結構體:
// Ready encapsulates the entries and messages that are ready to read,
// be saved to stable storage, committed or sent to other peers.
// All fields in Ready are read-only.
type Ready struct {
// The current volatile state of a Node.
// SoftState will be nil if there is no update.
// It is not required to consume or store SoftState.
*SoftState
// The current state of a Node to be saved to stable storage BEFORE
// Messages are sent.
// HardState will be equal to empty state if there is no update.
pb.HardState
// ReadStates can be used for node to serve linearizable read requests locally
// when its applied index is greater than the index in ReadState.
// Note that the readState will be returned when raft receives msgReadIndex.
// The returned is only valid for the request that requested to read.
ReadStates []ReadState
// Entries specifies entries to be saved to stable storage BEFORE
// Messages are sent.
Entries []pb.Entry
// Snapshot specifies the snapshot to be saved to stable storage.
Snapshot pb.Snapshot
// CommittedEntries specifies entries to be committed to a
// store/state-machine. These have previously been committed to stable
// store.
CommittedEntries []pb.Entry
// Messages specifies outbound messages to be sent AFTER Entries are
// committed to stable storage.
// If it contains a MsgSnap message, the application MUST report back to raft
// when the snapshot has been received or has failed by calling ReportSnapshot.
Messages []pb.Message
// MustSync indicates whether the HardState and Entries must be synchronously
// written to disk or if an asynchronous write is permissible.
MustSync bool
}
可以說,這個Ready結構體封裝了一批更新,這些更新包括:
- pb.HardState: 包含當前節點見過的最大的term,以及在這個term給誰投過票,已經當前節點知道的commit index
- Messages: 需要廣播給所有peers的消息
- CommittedEntries:已經commit了,還沒有apply到狀態機的日志
- Snapshot:需要持久化的快照
庫的使用者從node結構體提供的一個ready channel中不斷的pop出一個個的Ready進行處理,庫使用者通過如下方法拿到Ready channel:
func (n *node) Ready() <-chan Ready { return n.readyc }
應用需要對Ready的處理包括:
- 將HardState, Entries, Snapshot持久化到storage。
- 將Messages(上文提到的msgs)非阻塞的廣播給其他peers
- 將CommittedEntries(已經commit還沒有apply)應用到狀態機。
- 如果發現CommittedEntries中有成員變更類型的entry,調用node的ApplyConfChange()方法讓node知道(這里和raft論文不一樣,論文中只要節點收到了成員變更日志就應用)
- 調用Node.Advance()告訴raft node,這批狀態更新處理完了,狀態已經演進了,可以給我下一批Ready讓我處理。
應用通過raft.StartNode()來啟動raft中的一個副本,函數內部通過啟動一個goroutine運行
func (n *node) run(r *raft)
來啟動服務。
應用通過調用
func (n *node) Propose(ctx context.Context, data []byte) error
來Propose一個請求給raft,被raft開始處理后返回。
增刪節點通過調用
func (n *node) ProposeConfChange(ctx context.Context, cc pb.ConfChange) error
node結構體包含幾個重要的channel:
// node is the canonical implementation of the Node interface
type node struct {
propc chan pb.Message
recvc chan pb.Message
confc chan pb.ConfChange
confstatec chan pb.ConfState
readyc chan Ready
advancec chan struct{}
tickc chan struct{}
done chan struct{}
stop chan struct{}
status chan chan Status
logger Logger
}
-
propc: propc是一個沒有buffer的channel,應用通過Propose接口寫入的請求被封裝成Message被push到propc中,node的run方法從propc中pop出Message,append自己的raft log中,並且將Message放入mailbox中(raft結構體中的msgs []pb.Message),這個msgs會被封裝在Ready中,被應用從readyc中取出來,然后通過應用自定義的transport發送出去。
-
recvc: 應用自定義的transport在收到Message后需要調用
func (n *node) Step(ctx context.Context, m pb.Message) error來把Message放入recvc中,經過一些處理后,同樣,會把需要發送的Message放入到對應peers的mailbox中。后續通過自定義transport發送出去。
-
readyc/advancec: readyc和advancec都是沒有buffer的channel,node.run()內部把相關的一些狀態更新打包成Ready結構體(其中一種狀態就是上面提到的msgs)放入readyc中。應用從readyc中pop出Ready中,對相應的狀態進行處理,處理完成后,調用
rc.node.Advance()往advancec中push一個空結構體告訴raft,已經對這批Ready包含的狀態進行了相應的處理,node.run()內部從advancec中得到通知后,對內部一些狀態進行處理,比如把已經持久化到storage中的entries從內存(對應type unstable struct)中刪除等。
-
tickc:應用定期往tickc中push空結構體,node.run()會調用tick()函數,對於leader來說,tick()會給其他peers發心跳,對於follower來說,會檢查是否需要發起選主操作。
-
confc/confstatec:應用從Ready中拿出CommittedEntries,檢查其如果含有成員變更類型的日志,則需要調用
func (n *node) ApplyConfChange(cc pb.ConfChange) *pb.ConfState這個函數會push ConfChange到confc中,confc同樣是個無buffer的channel,node.run()內部會從confc中拿出ConfChange,然后進行真正的增減peers操作,之后將最新的成員組push到confstatec中,而ApplyConfChange函數從confstatec pop出最新的成員組返回給應用。
可以說,要想用上etcd的raft library還是需要了解不少東西的。
