draft.go

/*
 * Copyright (C) 2017 Dgraph Labs, Inc. and Contributors
 *
 * This program is free software: you can redistribute it and/or modify
 * it under the terms of the GNU Affero General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU Affero General Public License for more details.
 *
 * You should have received a copy of the GNU Affero General Public License
 * along with this program.  If not, see <http://www.gnu.org/licenses/>.
 */

package worker

import (
	"encoding/binary"
	"log"
	"math/rand"
	"sync"
	"time"

	"github.com/coreos/etcd/raft"
	"github.com/coreos/etcd/raft/raftpb"
	"golang.org/x/net/context"
	"golang.org/x/net/trace"

	"github.com/dgraph-io/badger/y"
	"github.com/dgraph-io/dgraph/conn"
	"github.com/dgraph-io/dgraph/posting"
	"github.com/dgraph-io/dgraph/protos"
	"github.com/dgraph-io/dgraph/raftwal"
	"github.com/dgraph-io/dgraph/schema"
	"github.com/dgraph-io/dgraph/x"
)

type proposalCtx struct {
	ch  chan error
	ctx context.Context
	cnt int // used for reference counting
	// Since each proposal consists of multiple tasks we need to store
	// non-nil error returned by task
	err   error
	index uint64
	n     *node
	// Used for writing all deltas at end
	txn *posting.Txn
}

type proposals struct {
	sync.RWMutex
	ids map[uint32]*proposalCtx
}

func (p *proposals) Store(pid uint32, pctx *proposalCtx) bool {
	p.Lock()
	defer p.Unlock()
	if _, has := p.ids[pid]; has {
		return false
	}
	p.ids[pid] = pctx
	return true
}

func (p *proposals) IncRef(pid uint32, count int) {
	p.Lock()
	defer p.Unlock()
	pd, has := p.ids[pid]
	x.AssertTrue(has)
	pd.cnt += count
	return
}

func (p *proposals) pctx(pid uint32) *proposalCtx {
	p.RLock()
	defer p.RUnlock()
	return p.ids[pid]
}

func (p *proposals) CtxAndTxn(pid uint32) (context.Context, *posting.Txn) {
	p.RLock()
	defer p.RUnlock()
	pd, has := p.ids[pid]
	x.AssertTrue(has)
	return pd.ctx, pd.txn
}

func (p *proposals) Done(pid uint32, err error) {
	p.Lock()
	defer p.Unlock()
	pd, has := p.ids[pid]
	if !has {
		return
	}
	x.AssertTrue(pd.cnt > 0 && pd.index != 0)
	pd.cnt -= 1
	if err != nil {
		pd.err = err
	}
	if pd.cnt > 0 {
		return
	}
	delete(p.ids, pid)
	pd.ch <- pd.err
	// We emit one pending watermark as soon as we read from rd.committedentries.
	// Since the tasks are executed in goroutines we need one guarding watermark which
	// is done only when all the pending sync/applied marks have been emitted.
	pd.n.Applied.Done(pd.index)
}

func (p *proposals) Has(pid uint32) bool {
	p.RLock()
	defer p.RUnlock()
	_, has := p.ids[pid]
	return has
}

type node struct {
	*conn.Node

	// Fields which are never changed after init.
	applyCh chan raftpb.Entry
	ctx     context.Context
	stop    chan struct{} // to send the stop signal to Run
	done    chan struct{} // to check whether node is running or not
	gid     uint32
	props   proposals

	canCampaign bool
	sch         *scheduler
}

func newNode(gid uint32, id uint64, myAddr string) *node {
	x.Printf("Node ID: %v with GroupID: %v\n", id, gid)

	rc := &protos.RaftContext{
		Addr:  myAddr,
		Group: gid,
		Id:    id,
	}
	m := conn.NewNode(rc)
	props := proposals{
		ids: make(map[uint32]*proposalCtx),
	}

	n := &node{
		Node: m,
		ctx:  context.Background(),
		gid:  gid,
		// processConfChange etc are not throttled so some extra delta, so that we don't
		// block tick when applyCh is full
		applyCh: make(chan raftpb.Entry, Config.NumPendingProposals+1000),
		props:   props,
		stop:    make(chan struct{}),
		done:    make(chan struct{}),
		sch:     new(scheduler),
	}
	n.sch.init(n)
	return n
}

type header struct {
	proposalId uint32
	msgId      uint16
}

func (h *header) Length() int {
	return 6 // 4 bytes for proposalId, 2 bytes for msgId.
}

func (h *header) Encode() []byte {
	result := make([]byte, h.Length())
	binary.LittleEndian.PutUint32(result[0:4], h.proposalId)
	binary.LittleEndian.PutUint16(result[4:6], h.msgId)
	return result
}

func (h *header) Decode(in []byte) {
	h.proposalId = binary.LittleEndian.Uint32(in[0:4])
	h.msgId = binary.LittleEndian.Uint16(in[4:6])
}

func (n *node) ProposeAndWait(ctx context.Context, proposal *protos.Proposal) error {
	if n.Raft() == nil {
		return x.Errorf("Raft isn't initialized yet")
	}
	// TODO: Should be based on number of edges (amount of work)
	pendingProposals <- struct{}{}
	x.PendingProposals.Add(1)
	defer func() { <-pendingProposals; x.PendingProposals.Add(-1) }()
	if ctx.Err() != nil {
		return ctx.Err()
	}
	// Do a type check here if schema is present
	// In very rare cases invalid entries might pass through raft, which would
	// be persisted, we do best effort schema check while writing
	if proposal.Mutations != nil {
		for _, edge := range proposal.Mutations.Edges {
			if tablet := groups().Tablet(edge.Attr); tablet != nil && tablet.ReadOnly {
				return errPredicateMoving
			}
			if typ, err := schema.State().TypeOf(edge.Attr); err != nil {
				continue
			} else if err := ValidateAndConvert(edge, typ); err != nil {
				return err
			}
		}
		for _, schema := range proposal.Mutations.Schema {
			if tablet := groups().Tablet(schema.Predicate); tablet != nil && tablet.ReadOnly {
				return errPredicateMoving
			}
			if err := checkSchema(schema); err != nil {
				return err
			}
		}
	}

	che := make(chan error, 1)
	pctx := &proposalCtx{
		ch:  che,
		ctx: ctx,
		n:   n,
		cnt: 1,
	}
	for {
		id := rand.Uint32()
		if n.props.Store(id, pctx) {
			proposal.Id = id
			break
		}
	}

	sz := proposal.Size()
	slice := make([]byte, sz)

	upto, err := proposal.MarshalTo(slice)
	if err != nil {
		return err
	}

	// We don't timeout on a mutation which has already been proposed.
	if err = n.Raft().Propose(ctx, slice[:upto]); err != nil {
		return x.Wrapf(err, "While proposing")
	}

	if tr, ok := trace.FromContext(ctx); ok {
		tr.LazyPrintf("Waiting for the proposal.")
	}

	err = <-che
	if err != nil {
		if tr, ok := trace.FromContext(ctx); ok {
			tr.LazyPrintf("Raft Propose error: %v", err)
		}
	}
	return err
}

func (n *node) processMutation(task *task) error {
	pid := task.pid
	ridx := task.rid
	edge := task.edge

	ctx, txn := n.props.CtxAndTxn(pid)
	if txn.ShouldAbort() {
		return posting.ErrConflict
	}
	rv := x.RaftValue{Group: n.gid, Index: ridx}
	ctx = context.WithValue(ctx, "raft", rv)
	if err := runMutation(ctx, edge, txn); err != nil {
		if tr, ok := trace.FromContext(ctx); ok {
			tr.LazyPrintf("process mutation: %v", err)
		}
		return err
	}
	return nil
}

func (n *node) processSchemaMutations(pid uint32, index uint64,
	startTs uint64, s *protos.SchemaUpdate) error {
	ctx, _ := n.props.CtxAndTxn(pid)
	rv := x.RaftValue{Group: n.gid, Index: index}
	ctx = context.WithValue(ctx, "raft", rv)
	if err := runSchemaMutation(ctx, s, startTs); err != nil {
		if tr, ok := trace.FromContext(n.ctx); ok {
			tr.LazyPrintf(err.Error())
		}
		return err
	}
	return nil
}

func (n *node) applyConfChange(e raftpb.Entry) {
	var cc raftpb.ConfChange
	cc.Unmarshal(e.Data)

	if cc.Type == raftpb.ConfChangeRemoveNode {
		n.DeletePeer(cc.NodeID)
	} else if len(cc.Context) > 0 {
		var rc protos.RaftContext
		x.Check(rc.Unmarshal(cc.Context))
		n.Connect(rc.Id, rc.Addr)
	}

	cs := n.Raft().ApplyConfChange(cc)
	n.SetConfState(cs)
	// Not present in proposal map
	n.Applied.Done(e.Index)
	groups().triggerMembershipSync()
}

func (n *node) processApplyCh() {
	for e := range n.applyCh {
		if len(e.Data) == 0 {
			// This is not in the proposal map
			n.Applied.Done(e.Index)
			continue
		}

		if e.Type == raftpb.EntryConfChange {
			n.applyConfChange(e)
			continue
		}

		x.AssertTrue(e.Type == raftpb.EntryNormal)

		proposal := &protos.Proposal{}
		if err := proposal.Unmarshal(e.Data); err != nil {
			log.Fatalf("Unable to unmarshal proposal: %v %q\n", err, e.Data)
		}

		// One final applied and synced watermark would be emitted when proposal ctx ref count
		// becomes zero.
		pctx := n.props.pctx(proposal.Id)
		if pctx == nil {
			// This is during replay of logs after restart
			pctx = &proposalCtx{
				ch:  make(chan error, 1),
				ctx: n.ctx,
				n:   n,
				cnt: 1,
			}
		}
		pctx.index = e.Index
		n.props.Store(proposal.Id, pctx)

		posting.TxnMarks().Begin(e.Index)
		if proposal.Mutations != nil {
			// syncmarks for this shouldn't be marked done until it's comitted.
			n.sch.schedule(proposal, e.Index)
		} else if len(proposal.Kv) > 0 {
			go n.processKeyValues(e.Index, proposal.Id, proposal.Kv)
		} else if proposal.State != nil {
			// This state needn't be snapshotted in this group, on restart we would fetch
			// a state which is latest or equal to this.
			groups().applyState(proposal.State)
			// When proposal is done it emits done watermarks.
			posting.TxnMarks().Done(e.Index)
			n.props.Done(proposal.Id, nil)
		} else if len(proposal.CleanPredicate) > 0 {
			go n.deletePredicate(e.Index, proposal.Id, proposal.CleanPredicate)
		} else if proposal.TxnContext != nil {
			go n.commitOrAbort(e.Index, proposal.Id, proposal.TxnContext)
		} else {
			x.Fatalf("Unknown proposal")
		}
	}
}

func (n *node) commitOrAbort(index uint64, pid uint32, tctx *protos.TxnContext) {
	ctx, _ := n.props.CtxAndTxn(pid)
	_, err := commitOrAbort(ctx, tctx)
	if tr, ok := trace.FromContext(ctx); ok {
		tr.LazyPrintf("Status of commitOrAbort %+v %v\n", tctx, err)
	}
	if err == nil {
		posting.Txns().Done(tctx.StartTs)
		posting.Oracle().Done(tctx.StartTs)
	}
	posting.TxnMarks().Done(index)
	n.props.Done(pid, err)
}

func (n *node) deletePredicate(index uint64, pid uint32, predicate string) {
	ctx, _ := n.props.CtxAndTxn(pid)
	rv := x.RaftValue{Group: n.gid, Index: index}
	ctx = context.WithValue(ctx, "raft", rv)
	err := posting.DeletePredicate(ctx, predicate)
	posting.TxnMarks().Done(index)
	n.props.Done(pid, err)
}

func (n *node) processKeyValues(index uint64, pid uint32, kvs []*protos.KV) error {
	ctx, _ := n.props.CtxAndTxn(pid)
	err := populateKeyValues(ctx, kvs)
	posting.TxnMarks().Done(index)
	n.props.Done(pid, err)
	return nil
}

func (n *node) applyAllMarks(ctx context.Context) {
	// Get index of last committed.
	lastIndex, err := n.Store.LastIndex()
	x.Checkf(err, "Error while getting last index")
	n.Applied.WaitForMark(ctx, lastIndex)
}

func (n *node) waitForTxnMarks(ctx context.Context) {
	// Get index of last committed.
	lastIndex, err := n.Store.LastIndex()
	x.Checkf(err, "Error while getting last index")
	posting.TxnMarks().WaitForMark(ctx, lastIndex)
}

func (n *node) retrieveSnapshot(peerID uint64) {
	addr, _ := n.Peer(peerID)
	pool, err := conn.Get().Get(addr)
	if err != nil {
		// err is just going to be errNoConnection
		log.Fatalf("Cannot retrieve snapshot from peer %v, no connection.  Error: %v\n",
			peerID, err)
	}

	// Wait for watermarks to sync since populateShard writes directly to db, otherwise
	// the values might get overwritten
	// Safe to keep this line
	n.applyAllMarks(n.ctx)
	// Need to clear pl's stored in memory for the case when retrieving snapshot with
	// index greater than this node's last index
	// Should invalidate/remove pl's to this group only ideally
	posting.EvictLRU()
	if _, err := populateShard(n.ctx, pstore, pool, n.gid); err != nil {
		// TODO: We definitely don't want to just fall flat on our face if we can't
		// retrieve a simple snapshot.
		log.Fatalf("Cannot retrieve snapshot from peer %v, error: %v\n", peerID, err)
	}
	// Populate shard stores the streamed data directly into db, so we need to refresh
	// schema for current group id
	x.Checkf(schema.LoadFromDb(), "Error while initilizating schema")
	groups().triggerMembershipSync()
}

func (n *node) Run() {
	firstRun := true
	var leader bool
	// See also our configuration of HeartbeatTick and ElectionTick.
	ticker := time.NewTicker(20 * time.Millisecond)
	defer ticker.Stop()
	rcBytes, err := n.RaftContext.Marshal()
	x.Check(err)

	// This chan could have capacity zero, because runReadIndexLoop never blocks without selecting
	// on readStateCh.  It's 2 so that sending rarely blocks (so the Go runtime doesn't have to
	// switch threads as much.)
	readStateCh := make(chan raft.ReadState, 2)
	closer := y.NewCloser(1)
	// We only stop runReadIndexLoop after the for loop below has finished interacting with it.
	// That way we know sending to readStateCh will not deadlock.
	defer closer.SignalAndWait()

	for {
		select {
		case <-ticker.C:
			n.Raft().Tick()

		case rd := <-n.Raft().Ready():
			for _, rs := range rd.ReadStates {
				readStateCh <- rs
			}

			if rd.SoftState != nil {
				groups().triggerMembershipSync()
				leader = rd.RaftState == raft.StateLeader
			}
			if leader {
				// Leader can send messages in parallel with writing to disk.
				for _, msg := range rd.Messages {
					// NOTE: We can do some optimizations here to drop messages.
					msg.Context = rcBytes
					n.Send(msg)
				}
			}

			// First store the entries, then the hardstate and snapshot.
			x.Check(n.Wal.Store(n.gid, rd.HardState, rd.Entries))
			x.Check(n.Wal.StoreSnapshot(n.gid, rd.Snapshot))

			// Now store them in the in-memory store.
			n.SaveToStorage(rd.Snapshot, rd.HardState, rd.Entries)

			if !raft.IsEmptySnap(rd.Snapshot) {
				// We don't send snapshots to other nodes. But, if we get one, that means
				// either the leader is trying to bring us up to state; or this is the
				// snapshot that I created. Only the former case should be handled.
				var rc protos.RaftContext
				x.Check(rc.Unmarshal(rd.Snapshot.Data))
				x.AssertTrue(rc.Group == n.gid)
				if rc.Id != n.Id {
					// NOTE: Retrieving snapshot here is OK, after storing it above in WAL, because
					// rc.Id != n.Id.
					x.Printf("-------> SNAPSHOT [%d] from %d\n", n.gid, rc.Id)
					// It's ok to block tick while retrieving snapshot, since it's a follower
					n.retrieveSnapshot(rc.Id)
					x.Printf("-------> SNAPSHOT [%d]. DONE.\n", n.gid)
				} else {
					x.Printf("-------> SNAPSHOT [%d] from %d [SELF]. Ignoring.\n", n.gid, rc.Id)
				}
			}
			if len(rd.CommittedEntries) > 0 {
				if tr, ok := trace.FromContext(n.ctx); ok {
					tr.LazyPrintf("Found %d committed entries", len(rd.CommittedEntries))
				}
			}

			// Now schedule or apply committed entries.
			for _, entry := range rd.CommittedEntries {
				// Need applied watermarks for schema mutation also for read linearazibility
				// Applied watermarks needs to be emitted as soon as possible sequentially.
				// If we emit Mark{4, false} and Mark{4, true} before emitting Mark{3, false}
				// then doneUntil would be set as 4 as soon as Mark{4,true} is done and before
				// Mark{3, false} is emitted. So it's safer to emit watermarks as soon as
				// possible sequentially
				n.Applied.Begin(entry.Index)

				if !leader && entry.Type == raftpb.EntryConfChange {
					// Config changes in followers must be applied straight away.
					n.applyConfChange(entry)
				} else {
					// TODO: Stop accepting requests when applyCh is full
					// Just queue up to be processed. Don't wait on them.
					n.applyCh <- entry
				}
			}

			if !leader {
				// Followers should send messages later.
				for _, msg := range rd.Messages {
					// NOTE: We can do some optimizations here to drop messages.
					msg.Context = rcBytes
					n.Send(msg)
				}
			}
			n.Raft().Advance()
			if firstRun && n.canCampaign {
				go n.Raft().Campaign(n.ctx)
				firstRun = false
			}

		case <-n.stop:
			if peerId, has := groups().MyPeer(); has && n.AmLeader() {
				n.Raft().TransferLeadership(n.ctx, Config.RaftId, peerId)
				go func() {
					select {
					case <-n.ctx.Done(): // time out
						if tr, ok := trace.FromContext(n.ctx); ok {
							tr.LazyPrintf("context timed out while transfering leadership")
						}
					case <-time.After(1 * time.Second):
						if tr, ok := trace.FromContext(n.ctx); ok {
							tr.LazyPrintf("Timed out transfering leadership")
						}
					}
					n.Raft().Stop()
					close(n.done)
				}()
			} else {
				n.Raft().Stop()
				close(n.done)
			}
		case <-n.done:
			return
		}
	}
}

func (n *node) Stop() {
	select {
	case n.stop <- struct{}{}:
	case <-n.done:
		// already stopped.
		return
	}
	<-n.done // wait for Run to respond.
}

func (n *node) snapshotPeriodically() {
	if n.gid == 0 {
		// Group zero is dedicated for membership information, whose state we don't persist.
		// So, taking snapshots would end up deleting the RAFT entries that we need to
		// regenerate the state on a crash. Therefore, don't take snapshots.
		return
	}

	ticker := time.NewTicker(time.Minute)
	defer ticker.Stop()

	for {
		select {
		case <-ticker.C:
			n.snapshot(Config.MaxPendingCount)

		case <-n.done:
			return
		}
	}
}

func (n *node) snapshot(skip uint64) {
	if n.gid == 0 {
		// Group zero is dedicated for membership information, whose state we don't persist.
		// So, taking snapshots would end up deleting the RAFT entries that we need to
		// regenerate the state on a crash. Therefore, don't take snapshots.
		return
	}
	water := posting.TxnMarks()
	le := water.DoneUntil()

	existing, err := n.Store.Snapshot()
	x.Checkf(err, "Unable to get existing snapshot")

	si := existing.Metadata.Index
	if le <= si+skip {
		return
	}
	snapshotIdx := le - skip
	if tr, ok := trace.FromContext(n.ctx); ok {
		tr.LazyPrintf("Taking snapshot for group: %d at watermark: %d\n", n.gid, snapshotIdx)
	}
	rc, err := n.RaftContext.Marshal()
	x.Check(err)

	s, err := n.Store.CreateSnapshot(snapshotIdx, n.ConfState(), rc)
	x.Checkf(err, "While creating snapshot")
	x.Checkf(n.Store.Compact(snapshotIdx), "While compacting snapshot")
	x.Check(n.Wal.StoreSnapshot(n.gid, s))
}

func (n *node) joinPeers() {
	// Get leader information for MY group.
	pl := groups().Leader(n.gid)
	if pl == nil {
		log.Fatalf("Unable to reach leader or any other server in group %d", n.gid)
	}

	// Bring the instance up to speed first.
	// Raft would decide whether snapshot needs to fetched or not
	// so populateShard is not needed
	// _, err := populateShard(n.ctx, pool, n.gid)
	// x.Checkf(err, "Error while populating shard")

	gconn := pl.Get()

	c := protos.NewRaftClient(gconn)
	x.Printf("Calling JoinCluster")
	_, err := c.JoinCluster(n.ctx, n.RaftContext)
	// TODO: This should keep on indefinitely trying to join the cluster, instead of crashing.
	x.Checkf(err, "Error while joining cluster")
	x.Printf("Done with JoinCluster call\n")
}

// InitAndStartNode gets called after having at least one membership sync with the cluster.
func (n *node) InitAndStartNode(wal *raftwal.Wal) {
	idx, restart, err := n.InitFromWal(wal)
	x.Check(err)
	n.Applied.SetDoneUntil(idx)
	posting.TxnMarks().SetDoneUntil(idx)

	if restart {
		x.Printf("Restarting node for group: %d\n", n.gid)
		found := groups().HasMeInState()
		_, hasPeer := groups().MyPeer()
		if !found && hasPeer {
			n.joinPeers()
		}
		n.SetRaft(raft.RestartNode(n.Cfg))

	} else {
		x.Printf("New Node for group: %d\n", n.gid)
		if _, hasPeer := groups().MyPeer(); hasPeer {
			n.joinPeers()
			n.SetRaft(raft.StartNode(n.Cfg, nil))

		} else {
			peers := []raft.Peer{{ID: n.Id}}
			n.SetRaft(raft.StartNode(n.Cfg, peers))
			// Trigger election, so this node can become the leader of this single-node cluster.
			n.canCampaign = true
		}
	}
	go n.processApplyCh()
	go n.Run()
	go n.snapshotPeriodically()
	go n.BatchAndSendMessages()
}

func (n *node) AmLeader() bool {
	if n.Raft() == nil {
		return false
	}
	r := n.Raft()
	return r.Status().Lead == r.Status().ID
}