[
https://issues.apache.org/jira/browse/RATIS-458?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16992878#comment-16992878
]
Tsz-wo Sze commented on RATIS-458:
----------------------------------
The approach looks great. Just one question:
- Should we timeout non-heart append requests? It seems not useful as
discussed previously.
Some other comments
- Let's add a isHeartbeat field to AppendEntriesReplyProto and
StatusRuntimeException. Then the leader side is more deterministic.
- Let's add a RequestMap class as below
{code}
//GrpcLogAppender
static class RequestMap {
private final Map<Long, AppendEntriesRequest> logRequests = new
ConcurrentHashMap<>();
private final Map<Long, AppendEntriesRequest> heartbeats = new
ConcurrentHashMap<>();
int logRequestsSize() {
return logRequests.size();
}
void clear() {
logRequests.clear();
heartbeats.clear();
}
void put(AppendEntriesRequest request) {
if (request.isHeartbeat()) {
heartbeats.put(request.getCallId(), request);
} else {
logRequests.put(request.getCallId(), request);
}
}
AppendEntriesRequest remove(AppendEntriesReplyProto reply) {
return remove(reply.getServerReply().getCallId(), reply.isHeartbeat());
}
AppendEntriesRequest remove(long cid, boolean isHeartbeat) {
return isHeartbeat? heartbeats.remove(cid): logRequests.remove(cid);
}
}
{code}
and use it as pendingRequests.
{code}
private final RequestMap pendingRequests = new RequestMap();
{code}
- Let move futures variable down and keep it final.
{code}
//RaftServerImpl.appendEntriesAsync
final List<CompletableFuture<Long>> futures = entries.length == 0?
Collections.emptyList()
: state.getLog().append(entries);
{code}
> GrpcLogAppender#shouldWait should wait on pending log entries to follower
> -------------------------------------------------------------------------
>
> Key: RATIS-458
> URL: https://issues.apache.org/jira/browse/RATIS-458
> Project: Ratis
> Issue Type: Improvement
> Components: gRPC
> Reporter: Lokesh Jain
> Assignee: Lokesh Jain
> Priority: Major
> Labels: ozone
> Attachments: RATIS-458.001.patch, RATIS-458.002.patch,
> RATIS-458.003.patch, RATIS-458.004.patch, RATIS-458.005.patch,
> RATIS-458.006.patch, RATIS-458.007.patch
>
>
> In GrpcLogAppender when an append entry times out we remove the entry from
> the pendingRequests. This decreases the size of pendingRequests which affects
> the logic in GrpcLogAppender#shouldWait. Further we also consider heartbeats
> in shouldWait because heartbeats are tracked in pendingRequests. It should
> actually wait on the number of log entries which are appended to follower but
> have not yet been processed by it.
> GrpcConfigKeys.Server.leaderOutstandingAppendsMax should also be a fraction
> of RaftServerConfigKeys.Log.queueSize. This brings flow control for leader's
> append entries to follower because then number of outstanding append entries
> in leader would be limited by maximum number of operations in raft log worker.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
