2016-05-29 04:22:57 +03:00
# Raft library
Raft is a protocol with which a cluster of nodes can maintain a replicated state machine.
The state machine is kept in sync through the use of a replicated log.
For more details on Raft, see "In Search of an Understandable Consensus Algorithm"
(https://ramcloud.stanford.edu/raft.pdf) by Diego Ongaro and John Ousterhout.
2016-07-29 20:55:11 +03:00
This Raft library is stable and feature complete. As of 2016, it is **the most widely used** Raft library in production, serving tens of thousands clusters each day. It powers distributed systems such as etcd, Kubernetes, Docker Swarm, Cloud Foundry Diego, CockroachDB, TiDB, Project Calico, Flannel, and more.
Most Raft implementations have a monolithic design, including storage handling, messaging serialization, and network transport. This library instead follows a minimalistic design philosophy by only implementing the core raft algorithm. This minimalism buys flexibility, determinism, and performance.
To keep the codebase small as well as provide flexibility, the library only implements the Raft algorithm; both network and disk IO are left to the user. Library users must implement their own transportation layer for message passing between Raft peers over the wire. Similarly, users must implement their own storage layer to persist the Raft log and state.
In order to easily test the Raft library, its behavior should be deterministic. To achieve this determinism, the library models Raft as a state machine. The state machine takes a `Message` as input. A message can either be a local timer update or a network message sent from a remote peer. The state machine's output is a 3-tuple `{[]Messages, []LogEntries, NextState}` consisting of an array of `Messages` , `log entries` , and `Raft state changes` . For state machines with the same state, the same state machine input should always generate the same state machine output.
2016-05-29 04:22:57 +03:00
A simple example application, _raftexample_ , is also available to help illustrate
how to use this package in practice:
https://github.com/coreos/etcd/tree/master/contrib/raftexample
2016-07-02 21:03:57 +03:00
# Features
This raft implementation is a full feature implementation of Raft protocol. Features includes:
- Leader election
- Log replication
- Log compaction
- Membership changes
- Leadership transfer extension
- Lease-based linearizable read-only queries served by both the leader and followers
This raft implementation also includes a few optional enhancements:
- Optimistic pipelining to reduce log replication latency
- Flow control for log replication
- Batching Raft messages to reduce synchronized network I/O calls
- Batching log entries to reduce disk synchronized I/O
- Writing to leader's disk in parallel
- Internal proposal redirection from followers to leader
- Automatic stepping down when the leader loses quorum
2016-06-09 20:10:44 +03:00
## Notable Users
2016-05-29 04:22:57 +03:00
- [cockroachdb ](https://github.com/cockroachdb/cockroach ) A Scalable, Survivable, Strongly-Consistent SQL Database
- [etcd ](https://github.com/coreos/etcd ) A distributed reliable key-value store
- [tikv ](https://github.com/pingcap/tikv ) Distributed transactional key value database powered by Rust and Raft
2016-06-09 20:10:44 +03:00
- [swarmkit ](https://github.com/docker/swarmkit ) A toolkit for orchestrating distributed systems at any scale.
2016-05-29 04:22:57 +03:00
## Usage
The primary object in raft is a Node. You either start a Node from scratch
using raft.StartNode or start a Node from some initial state using raft.RestartNode.
To start a node from scratch:
```go
storage := raft.NewMemoryStorage()
c := & Config{
ID: 0x01,
ElectionTick: 10,
HeartbeatTick: 1,
Storage: storage,
MaxSizePerMsg: 4096,
MaxInflightMsgs: 256,
}
n := raft.StartNode(c, []raft.Peer{{ID: 0x02}, {ID: 0x03}})
```
To restart a node from previous state:
```go
storage := raft.NewMemoryStorage()
// recover the in-memory storage from persistent
// snapshot, state and entries.
storage.ApplySnapshot(snapshot)
storage.SetHardState(state)
storage.Append(entries)
c := & Config{
ID: 0x01,
ElectionTick: 10,
HeartbeatTick: 1,
Storage: storage,
MaxSizePerMsg: 4096,
MaxInflightMsgs: 256,
}
// restart raft without peer information.
// peer information is already included in the storage.
n := raft.RestartNode(c)
```
Now that you are holding onto a Node you have a few responsibilities:
First, you must read from the Node.Ready() channel and process the updates
it contains. These steps may be performed in parallel, except as noted in step
2.
1. Write HardState, Entries, and Snapshot to persistent storage if they are
not empty. Note that when writing an Entry with Index i, any
previously-persisted entries with Index >= i must be discarded.
2. Send all Messages to the nodes named in the To field. It is important that
no messages be sent until the latest HardState has been persisted to disk,
and all Entries written by any previous Ready batch (Messages may be sent while
entries from the same batch are being persisted). To reduce the I/O latency, an
optimization can be applied to make leader write to disk in parallel with its
followers (as explained at section 10.2.1 in Raft thesis). If any Message has type
MsgSnap, call Node.ReportSnapshot() after it has been sent (these messages may be
large). Note: Marshalling messages is not thread-safe; it is important that you
make sure that no new entries are persisted while marshalling.
The easiest way to achieve this is to serialise the messages directly inside
your main raft loop.
3. Apply Snapshot (if any) and CommittedEntries to the state machine.
If any committed Entry has Type EntryConfChange, call Node.ApplyConfChange()
to apply it to the node. The configuration change may be cancelled at this point
by setting the NodeID field to zero before calling ApplyConfChange
(but ApplyConfChange must be called one way or the other, and the decision to cancel
must be based solely on the state machine and not external information such as
the observed health of the node).
4. Call Node.Advance() to signal readiness for the next batch of updates.
This may be done at any time after step 1, although all updates must be processed
in the order they were returned by Ready.
Second, all persisted log entries must be made available via an
implementation of the Storage interface. The provided MemoryStorage
type can be used for this (if you repopulate its state upon a
restart), or you can supply your own disk-backed implementation.
Third, when you receive a message from another node, pass it to Node.Step:
```go
func recvRaftRPC(ctx context.Context, m raftpb.Message) {
n.Step(ctx, m)
}
```
Finally, you need to call `Node.Tick()` at regular intervals (probably
via a `time.Ticker` ). Raft has two important timeouts: heartbeat and the
election timeout. However, internally to the raft package time is
represented by an abstract "tick".
The total state machine handling loop will look something like this:
```go
for {
select {
case < -s.Ticker:
n.Tick()
case rd := < -s.Node.Ready ( ) :
saveToStorage(rd.State, rd.Entries, rd.Snapshot)
send(rd.Messages)
if !raft.IsEmptySnap(rd.Snapshot) {
processSnapshot(rd.Snapshot)
}
for _, entry := range rd.CommittedEntries {
process(entry)
if entry.Type == raftpb.EntryConfChange {
var cc raftpb.ConfChange
cc.Unmarshal(entry.Data)
s.Node.ApplyConfChange(cc)
}
}
s.Node.Advance()
case < -s.done:
return
}
}
```
To propose changes to the state machine from your node take your application
data, serialize it into a byte slice and call:
```go
n.Propose(ctx, data)
```
If the proposal is committed, data will appear in committed entries with type
raftpb.EntryNormal. There is no guarantee that a proposed command will be
committed; you may have to re-propose after a timeout.
To add or remove node in a cluster, build ConfChange struct 'cc' and call:
```go
n.ProposeConfChange(ctx, cc)
```
After config change is committed, some committed entry with type
raftpb.EntryConfChange will be returned. You must apply it to node through:
```go
var cc raftpb.ConfChange
cc.Unmarshal(data)
n.ApplyConfChange(cc)
```
Note: An ID represents a unique node in a cluster for all time. A
given ID MUST be used only once even if the old node has been removed.
This means that for example IP addresses make poor node IDs since they
may be reused. Node IDs must be non-zero.
## Implementation notes
This implementation is up to date with the final Raft thesis
(https://ramcloud.stanford.edu/~ongaro/thesis.pdf), although our
implementation of the membership change protocol differs somewhat from
that described in chapter 4. The key invariant that membership changes
happen one node at a time is preserved, but in our implementation the
membership change takes effect when its entry is applied, not when it
is added to the log (so the entry is committed under the old
membership instead of the new). This is equivalent in terms of safety,
since the old and new configurations are guaranteed to overlap.
To ensure that we do not attempt to commit two membership changes at
once by matching log positions (which would be unsafe since they
should have different quorum requirements), we simply disallow any
proposed membership change while any uncommitted change appears in
the leader's log.
This approach introduces a problem when you try to remove a member
from a two-member cluster: If one of the members dies before the
other one receives the commit of the confchange entry, then the member
cannot be removed any more since the cluster cannot make progress.
For this reason it is highly recommended to use three or more nodes in
every cluster.