func stepFollower(r *raft, m pb.Message) {
switch m.Type {
case pb.MsgProp:
if r.lead == None {
r.logger.Infof("%x no leader at term %d; dropping proposal", r.id, r.Term)
return
}
m.To = r.lead
r.send(m)
case pb.MsgApp:
r.electionElapsed = 0
r.lead = m.From
r.handleAppendEntries(m)
case pb.MsgHeartbeat:
r.electionElapsed = 0
r.lead = m.From
r.handleHeartbeat(m)
case pb.MsgSnap:
r.electionElapsed = 0
r.lead = m.From
r.handleSnapshot(m)
case pb.MsgVote:
if (r.Vote == None || r.Vote == m.From) && r.raftLog.isUpToDate(m.Index, m.LogTerm) {
r.electionElapsed = 0
r.logger.Infof("%x [logterm: %d, index: %d, vote: %x] voted for %x [logterm: %d, index: %d] at term %d",
r.id, r.raftLog.lastTerm(), r.raftLog.lastIndex(), r.Vote, m.From, m.LogTerm, m.Index, r.Term)
r.Vote = m.From
r.send(pb.Message{To: m.From, Type: pb.MsgVoteResp})
} else {
r.logger.Infof("%x [logterm: %d, index: %d, vote: %x] rejected vote from %x [logterm: %d, index: %d] at term %d",
r.id, r.raftLog.lastTerm(), r.raftLog.lastIndex(), r.Vote, m.From, m.LogTerm, m.Index, r.Term)
r.send(pb.Message{To: m.From, Type: pb.MsgVoteResp, Reject: true})
}
case pb.MsgTransferLeader:
if r.lead == None {
r.logger.Infof("%x no leader at term %d; dropping leader transfer msg", r.id, r.Term)
return
}
m.To = r.lead
r.send(m)
case pb.MsgTimeoutNow:
r.logger.Infof("%x [term %d] received MsgTimeoutNow from %x and starts an election to get leadership.", r.id, r.Term, m.From)
r.campaign(campaignTransfer)
case pb.MsgReadIndex:
if r.lead == None {
r.logger.Infof("%x no leader at term %d; dropping index reading msg", r.id, r.Term)
return
}
m.To = r.lead
r.send(m)
case pb.MsgReadIndexResp:
if len(m.Entries) != 1 {
r.logger.Errorf("%x invalid format of MsgReadIndexResp from %x, entries count: %d", r.id, m.From, len(m.Entries))
return
}
r.readState.Index = m.Index
r.readState.RequestCtx = m.Entries[0].Data
}
}
func stepFollower(r *raft, m pb.Message) {
switch m.Type {
case pb.MsgProp:
if r.lead == None {
r.logger.Infof("%x no leader at term %d; dropping proposal", r.id, r.Term)
return
}
m.To = r.lead
r.send(m)
case pb.MsgApp:
r.electionElapsed = 0
r.lead = m.From
r.handleAppendEntries(m)
case pb.MsgHeartbeat:
r.electionElapsed = 0
r.lead = m.From
r.handleHeartbeat(m)
case pb.MsgSnap:
r.electionElapsed = 0
r.lead = m.From
r.handleSnapshot(m)
case pb.MsgTransferLeader:
if r.lead == None {
r.logger.Infof("%x no leader at term %d; dropping leader transfer msg", r.id, r.Term)
return
}
m.To = r.lead
r.send(m)
case pb.MsgTimeoutNow:
if r.promotable() {
r.logger.Infof("%x [term %d] received MsgTimeoutNow from %x and starts an election to get leadership.", r.id, r.Term, m.From)
// Leadership transfers never use pre-vote even if r.preVote is true; we
// know we are not recovering from a partition so there is no need for the
// extra round trip.
r.campaign(campaignTransfer)
} else {
r.logger.Infof("%x received MsgTimeoutNow from %x but is not promotable", r.id, m.From)
}
case pb.MsgReadIndex:
if r.lead == None {
r.logger.Infof("%x no leader at term %d; dropping index reading msg", r.id, r.Term)
return
}
m.To = r.lead
r.send(m)
case pb.MsgReadIndexResp:
if len(m.Entries) != 1 {
r.logger.Errorf("%x invalid format of MsgReadIndexResp from %x, entries count: %d", r.id, m.From, len(m.Entries))
return
}
r.readStates = append(r.readStates, ReadState{Index: m.Index, RequestCtx: m.Entries[0].Data})
}
}
func stepFollower(r *raft, m pb.Message) {
switch m.Type {
case pb.MsgProp:
if r.lead == None {
r.logger.Infof("%x no leader at term %d; dropping proposal", r.id, r.Term)
return
}
m.To = r.lead
r.send(m)
case pb.MsgApp:
r.electionElapsed = 0
r.lead = m.From
r.handleAppendEntries(m)
case pb.MsgHeartbeat:
r.electionElapsed = 0
r.lead = m.From
r.handleHeartbeat(m)
case pb.MsgSnap:
r.electionElapsed = 0
r.handleSnapshot(m)
case pb.MsgVote:
if (r.Vote == None || r.Vote == m.From) && r.raftLog.isUpToDate(m.Index, m.LogTerm) {
r.electionElapsed = 0
r.logger.Infof("%x [logterm: %d, index: %d, vote: %x] voted for %x [logterm: %d, index: %d] at term %d",
r.id, r.raftLog.lastTerm(), r.raftLog.lastIndex(), r.Vote, m.From, m.LogTerm, m.Index, r.Term)
r.Vote = m.From
r.send(pb.Message{To: m.From, Type: pb.MsgVoteResp})
} else {
r.logger.Infof("%x [logterm: %d, index: %d, vote: %x] rejected vote from %x [logterm: %d, index: %d] at term %d",
r.id, r.raftLog.lastTerm(), r.raftLog.lastIndex(), r.Vote, m.From, m.LogTerm, m.Index, r.Term)
r.send(pb.Message{To: m.From, Type: pb.MsgVoteResp, Reject: true})
}
}
}
func stepFollower(r *raft, m pb.Message) {
switch m.Type {
case msgProp:
if r.lead == None {
panic("no leader")
}
m.To = r.lead
r.send(m)
case msgApp:
r.elapsed = 0
r.lead = m.From
r.handleAppendEntries(m)
case msgSnap:
r.elapsed = 0
r.handleSnapshot(m)
case msgVote:
if (r.Vote == None || r.Vote == m.From) && r.raftLog.isUpToDate(m.Index, m.LogTerm) {
r.elapsed = 0
r.Vote = m.From
r.send(pb.Message{To: m.From, Type: msgVoteResp, Index: r.raftLog.lastIndex()})
} else {
r.send(pb.Message{To: m.From, Type: msgVoteResp, Index: -1})
}
}
}
// sendAppend sends RRPC, with entries to the given peer.
func (r *raft) sendAppend(to uint64) {
pr := r.prs[to]
m := pb.Message{}
m.To = to
if r.needSnapshot(pr.next) {
m.Type = pb.MsgSnap
snapshot, err := r.raftLog.snapshot()
if err != nil {
panic(err) // TODO(bdarnell)
}
if IsEmptySnap(snapshot) {
panic("need non-empty snapshot")
}
m.Snapshot = snapshot
sindex, sterm := snapshot.Metadata.Index, snapshot.Metadata.Term
log.Printf("raft: %x [firstindex: %d, commit: %d] sent snapshot[index: %d, term: %d] to %x [%s]",
r.id, r.raftLog.firstIndex(), r.Commit, sindex, sterm, to, pr)
} else {
m.Type = pb.MsgApp
m.Index = pr.next - 1
m.LogTerm = r.raftLog.term(pr.next - 1)
m.Entries = r.raftLog.entries(pr.next)
m.Commit = r.raftLog.committed
// optimistically increase the next if the follower
// has been matched.
if n := len(m.Entries); pr.match != 0 && n != 0 {
pr.optimisticUpdate(m.Entries[n-1].Index)
}
}
r.send(m)
}
func stepFollower(r *raft, m pb.Message) {
switch m.Type {
case pb.MsgProp:
if r.lead == None {
panic("no leader")
}
m.To = r.lead
r.send(m)
case pb.MsgApp:
r.elapsed = 0
r.lead = m.From
if m.LogTerm == 0 && m.Index == 0 && len(m.Entries) == 0 {
r.handleHeartbeat(m)
} else {
r.handleAppendEntries(m)
}
case pb.MsgSnap:
r.elapsed = 0
r.handleSnapshot(m)
case pb.MsgVote:
if (r.Vote == None || r.Vote == m.From) && r.raftLog.isUpToDate(m.Index, m.LogTerm) {
r.elapsed = 0
r.Vote = m.From
r.send(pb.Message{To: m.From, Type: pb.MsgVoteResp})
} else {
r.send(pb.Message{To: m.From, Type: pb.MsgVoteResp, Reject: true})
}
}
}
// sendAppend sends RRPC, with entries to the given peer.
func (r *raft) sendAppend(to uint64) {
pr := r.prs[to]
m := pb.Message{}
m.To = to
if r.needSnapshot(pr.next) {
m.Type = pb.MsgSnap
snapshot, err := r.raftLog.snapshot()
if err != nil {
panic(err) // TODO(bdarnell)
}
if IsEmptySnap(snapshot) {
panic("need non-empty snapshot")
}
m.Snapshot = snapshot
} else {
m.Type = pb.MsgApp
m.Index = pr.next - 1
m.LogTerm = r.raftLog.term(pr.next - 1)
m.Entries = r.raftLog.entries(pr.next)
m.Commit = r.raftLog.committed
// optimistically increase the next if the follower
// has been matched.
if n := len(m.Entries); pr.match != 0 && n != 0 {
pr.optimisticUpdate(m.Entries[n-1].Index)
}
}
r.send(m)
}
// sendAppend sends RPC, with entries to the given peer.
func (r *raft) sendAppend(to uint64) {
pr := r.prs[to]
if pr.isPaused() {
return
}
m := pb.Message{}
m.To = to
term, errt := r.raftLog.term(pr.Next - 1)
ents, erre := r.raftLog.entries(pr.Next, r.maxMsgSize)
if errt != nil || erre != nil { // send snapshot if we failed to get term or entries
if !pr.RecentActive {
r.logger.Debugf("ignore sending snapshot to %x since it is not recently active", to)
return
}
m.Type = pb.MsgSnap
snapshot, err := r.raftLog.snapshot()
if err != nil {
if err == ErrSnapshotTemporarilyUnavailable {
r.logger.Debugf("%x failed to send snapshot to %x because snapshot is temporarily unavailable", r.id, to)
return
}
panic(err) // TODO(bdarnell)
}
if IsEmptySnap(snapshot) {
panic("need non-empty snapshot")
}
m.Snapshot = snapshot
sindex, sterm := snapshot.Metadata.Index, snapshot.Metadata.Term
r.logger.Debugf("%x [firstindex: %d, commit: %d] sent snapshot[index: %d, term: %d] to %x [%s]",
r.id, r.raftLog.firstIndex(), r.raftLog.committed, sindex, sterm, to, pr)
pr.becomeSnapshot(sindex)
r.logger.Debugf("%x paused sending replication messages to %x [%s]", r.id, to, pr)
} else {
m.Type = pb.MsgApp
m.Index = pr.Next - 1
m.LogTerm = term
m.Entries = ents
m.Commit = r.raftLog.committed
if n := len(m.Entries); n != 0 {
switch pr.State {
// optimistically increase the next when in ProgressStateReplicate
case ProgressStateReplicate:
last := m.Entries[n-1].Index
pr.optimisticUpdate(last)
pr.ins.add(last)
case ProgressStateProbe:
pr.pause()
default:
r.logger.Panicf("%x is sending append in unhandled state %s", r.id, pr.State)
}
}
}
r.send(m)
}
// Send a message. Returns false if the message was dropped.
func (rttc *raftTransportTestContext) Send(
from, to roachpb.ReplicaDescriptor, rangeID roachpb.RangeID, msg raftpb.Message,
) bool {
msg.To = uint64(to.ReplicaID)
msg.From = uint64(from.ReplicaID)
req := &storage.RaftMessageRequest{
RangeID: rangeID,
Message: msg,
ToReplica: to,
FromReplica: from,
}
return rttc.transports[from.NodeID].SendAsync(req)
}
// sendAppend sends RRPC, with entries to the given peer.
func (r *raft) sendAppend(to int64) {
pr := r.prs[to]
m := pb.Message{}
m.To = to
m.Index = pr.next - 1
if r.needSnapshot(m.Index) {
m.Type = msgSnap
m.Snapshot = r.raftLog.snapshot
} else {
m.Type = msgApp
m.LogTerm = r.raftLog.term(pr.next - 1)
m.Entries = r.raftLog.entries(pr.next)
m.Commit = r.raftLog.committed
}
r.send(m)
}
// sendAppend sends RRPC, with entries to the given peer.
func (r *raft) sendAppend(to uint64) {
pr := r.prs[to]
if pr.isPaused() {
return
}
m := pb.Message{}
m.To = to
if r.needSnapshot(pr.Next) {
m.Type = pb.MsgSnap
snapshot, err := r.raftLog.snapshot()
if err != nil {
panic(err) // TODO(bdarnell)
}
if IsEmptySnap(snapshot) {
panic("need non-empty snapshot")
}
m.Snapshot = snapshot
sindex, sterm := snapshot.Metadata.Index, snapshot.Metadata.Term
raftLogger.Infof("%x [firstindex: %d, commit: %d] sent snapshot[index: %d, term: %d] to %x [%s]",
r.id, r.raftLog.firstIndex(), r.Commit, sindex, sterm, to, pr)
pr.becomeSnapshot(sindex)
raftLogger.Infof("%x paused sending replication messages to %x [%s]", r.id, to, pr)
} else {
m.Type = pb.MsgApp
m.Index = pr.Next - 1
m.LogTerm = r.raftLog.term(pr.Next - 1)
m.Entries = r.raftLog.entries(pr.Next, r.maxMsgSize)
m.Commit = r.raftLog.committed
if n := len(m.Entries); n != 0 {
switch pr.State {
// optimistically increase the next when in ProgressStateReplicate
case ProgressStateReplicate:
last := m.Entries[n-1].Index
pr.optimisticUpdate(last)
pr.ins.add(last)
case ProgressStateProbe:
pr.pause()
default:
raftLogger.Panicf("%x is sending append in unhandled state %s", r.id, pr.State)
}
}
}
r.send(m)
}
// Send a message. Returns false if the message was dropped.
func (rttc *raftTransportTestContext) Send(
from, to roachpb.ReplicaDescriptor, rangeID roachpb.RangeID, msg raftpb.Message,
) bool {
msg.To = uint64(to.ReplicaID)
msg.From = uint64(from.ReplicaID)
req := &storage.RaftMessageRequest{
RangeID: rangeID,
Message: msg,
ToReplica: to,
FromReplica: from,
}
sender := rttc.transports[from.NodeID].MakeSender(
func(err error, _ roachpb.ReplicaDescriptor) {
if err != nil && !grpcutil.IsClosedConnection(err) &&
!testutils.IsError(err, channelServerBrokenRangeMessage) {
rttc.t.Fatal(err)
}
})
return sender.SendAsync(req)
}
// sendAppend sends RRPC, with entries to the given peer.
func (r *raft) sendAppend(to uint64) {
pr := r.prs[to]
if pr.shouldWait() {
log.Printf("raft: %x ignored sending %s to %x [%s]", r.id, pb.MsgApp, to, pr)
return
}
m := pb.Message{}
m.To = to
if r.needSnapshot(pr.next) {
m.Type = pb.MsgSnap
snapshot, err := r.raftLog.snapshot()
if err != nil {
panic(err) // TODO(bdarnell)
}
if IsEmptySnap(snapshot) {
panic("need non-empty snapshot")
}
m.Snapshot = snapshot
sindex, sterm := snapshot.Metadata.Index, snapshot.Metadata.Term
log.Printf("raft: %x [firstindex: %d, commit: %d] sent snapshot[index: %d, term: %d] to %x [%s]",
r.id, r.raftLog.firstIndex(), r.Commit, sindex, sterm, to, pr)
pr.waitSet(r.electionTimeout)
} else {
m.Type = pb.MsgApp
m.Index = pr.next - 1
m.LogTerm = r.raftLog.term(pr.next - 1)
m.Entries = r.raftLog.entries(pr.next)
m.Commit = r.raftLog.committed
// optimistically increase the next if the follower
// has been matched.
if n := len(m.Entries); pr.match != 0 && n != 0 {
pr.optimisticUpdate(m.Entries[n-1].Index)
} else if pr.match == 0 {
// TODO (xiangli): better way to find out if the follower is in good path or not
// a follower might be in bad path even if match != 0, since we optimistically
// increase the next.
pr.waitSet(r.heartbeatTimeout)
}
}
r.send(m)
}
