Beispiel #1
0
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in C#:
//ORIGINAL LINE: static void appendNewEntry(org.neo4j.causalclustering.core.consensus.state.ReadableRaftState ctx, org.neo4j.causalclustering.core.consensus.outcome.Outcome outcome, org.neo4j.causalclustering.core.replication.ReplicatedContent content) throws java.io.IOException
        internal static void AppendNewEntry(ReadableRaftState ctx, Outcome outcome, ReplicatedContent content)
        {
            long prevLogIndex = ctx.EntryLog().appendIndex();
            long prevLogTerm  = prevLogIndex == -1 ? -1 : prevLogIndex > ctx.LastLogIndexBeforeWeBecameLeader() ? ctx.Term() : ctx.EntryLog().readEntryTerm(prevLogIndex);

            RaftLogEntry newLogEntry = new RaftLogEntry(ctx.Term(), content);

            outcome.AddShipCommand(new ShipCommand.NewEntries(prevLogIndex, prevLogTerm, new RaftLogEntry[] { newLogEntry }));
            outcome.AddLogCommand(new AppendLogEntry(prevLogIndex + 1, newLogEntry));
        }
Beispiel #2
0
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in C#:
//ORIGINAL LINE: static void appendNewEntries(org.neo4j.causalclustering.core.consensus.state.ReadableRaftState ctx, org.neo4j.causalclustering.core.consensus.outcome.Outcome outcome, java.util.Collection<org.neo4j.causalclustering.core.replication.ReplicatedContent> contents) throws java.io.IOException
        internal static void AppendNewEntries(ReadableRaftState ctx, Outcome outcome, ICollection <ReplicatedContent> contents)
        {
            long prevLogIndex = ctx.EntryLog().appendIndex();
            long prevLogTerm  = prevLogIndex == -1 ? -1 : prevLogIndex > ctx.LastLogIndexBeforeWeBecameLeader() ? ctx.Term() : ctx.EntryLog().readEntryTerm(prevLogIndex);

//JAVA TO C# CONVERTER TODO TASK: Method reference constructor syntax is not converted by Java to C# Converter:
            RaftLogEntry[] raftLogEntries = contents.Select(content => new RaftLogEntry(ctx.Term(), content)).ToArray(RaftLogEntry[] ::new);

            outcome.AddShipCommand(new ShipCommand.NewEntries(prevLogIndex, prevLogTerm, raftLogEntries));
            outcome.AddLogCommand(new BatchAppendLogEntries(prevLogIndex + 1, 0, raftLogEntries));
        }
Beispiel #3
0
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in C#:
//ORIGINAL LINE: static void handleAppendEntriesRequest(org.neo4j.causalclustering.core.consensus.state.ReadableRaftState state, org.neo4j.causalclustering.core.consensus.outcome.Outcome outcome, org.neo4j.causalclustering.core.consensus.RaftMessages_AppendEntries_Request request, org.neo4j.logging.Log log) throws java.io.IOException
        internal static void HandleAppendEntriesRequest(ReadableRaftState state, Outcome outcome, Org.Neo4j.causalclustering.core.consensus.RaftMessages_AppendEntries_Request request, Log log)
        {
            if (request.LeaderTerm() < state.Term())
            {
                Org.Neo4j.causalclustering.core.consensus.RaftMessages_AppendEntries_Response appendResponse = new Org.Neo4j.causalclustering.core.consensus.RaftMessages_AppendEntries_Response(state.Myself(), state.Term(), false, -1, state.EntryLog().appendIndex());

                outcome.AddOutgoingMessage(new Org.Neo4j.causalclustering.core.consensus.RaftMessages_Directed(request.From(), appendResponse));
                return;
            }

            outcome.PreElection  = false;
            outcome.NextTerm     = request.LeaderTerm();
            outcome.Leader       = request.From();
            outcome.LeaderCommit = request.LeaderCommit();

            if (!Follower.LogHistoryMatches(state, request.PrevLogIndex(), request.PrevLogTerm()))
            {
                Debug.Assert(request.PrevLogIndex() > -1 && request.PrevLogTerm() > -1);
                Org.Neo4j.causalclustering.core.consensus.RaftMessages_AppendEntries_Response appendResponse = new Org.Neo4j.causalclustering.core.consensus.RaftMessages_AppendEntries_Response(state.Myself(), request.LeaderTerm(), false, -1, state.EntryLog().appendIndex());

                outcome.AddOutgoingMessage(new Org.Neo4j.causalclustering.core.consensus.RaftMessages_Directed(request.From(), appendResponse));
                return;
            }

            long baseIndex = request.PrevLogIndex() + 1;
            int  offset;

            /* Find possible truncation point. */
            for (offset = 0; offset < request.Entries().Length; offset++)
            {
                long logIndex = baseIndex + offset;
                long logTerm  = state.EntryLog().readEntryTerm(logIndex);

                if (logIndex > state.EntryLog().appendIndex())
                {
                    // entry doesn't exist because it's beyond the current log end, so we can go ahead and append
                    break;
                }
                else if (logIndex < state.EntryLog().prevIndex())
                {
                    // entry doesn't exist because it's before the earliest known entry, so continue with the next one
                    continue;
                }
                else if (logTerm != request.Entries()[offset].term())
                {
                    /*
                     * the entry's index falls within our current range and the term doesn't match what we know. We must
                     * truncate.
                     */
                    if (logIndex <= state.CommitIndex())                                // first, assert that we haven't committed what we are about to truncate
                    {
                        throw new System.InvalidOperationException(format("Cannot truncate entry at index %d with term %d when commit index is at %d", logIndex, logTerm, state.CommitIndex()));
                    }
                    outcome.AddLogCommand(new TruncateLogCommand(logIndex));
                    break;
                }
            }

            if (offset < request.Entries().Length)
            {
                outcome.AddLogCommand(new BatchAppendLogEntries(baseIndex, offset, request.Entries()));
            }

            Follower.CommitToLogOnUpdate(state, request.PrevLogIndex() + request.Entries().Length, request.LeaderCommit(), outcome);

            long endMatchIndex = request.PrevLogIndex() + request.Entries().Length;               // this is the index of the last incoming entry

            Org.Neo4j.causalclustering.core.consensus.RaftMessages_AppendEntries_Response appendResponse = new Org.Neo4j.causalclustering.core.consensus.RaftMessages_AppendEntries_Response(state.Myself(), request.LeaderTerm(), true, endMatchIndex, endMatchIndex);
            outcome.AddOutgoingMessage(new Org.Neo4j.causalclustering.core.consensus.RaftMessages_Directed(request.From(), appendResponse));
        }