cassandra-pr mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From aweisberg <...@git.apache.org>
Subject [GitHub] cassandra pull request #224: 14405 replicas
Date Wed, 16 May 2018 17:06:05 GMT
Github user aweisberg commented on a diff in the pull request:

    https://github.com/apache/cassandra/pull/224#discussion_r187442385
  
    --- Diff: src/java/org/apache/cassandra/service/reads/DataResolver.java ---
    @@ -64,12 +72,19 @@ public PartitionIterator resolve()
             // at the beginning of this method), so grab the response count once and use
that through the method.
             int count = responses.size();
             List<UnfilteredPartitionIterator> iters = new ArrayList<>(count);
    -        InetAddressAndPort[] sources = new InetAddressAndPort[count];
    +        Replica[] sources = new Replica[count];
             for (int i = 0; i < count; i++)
             {
                 MessageIn<ReadResponse> msg = responses.get(i);
                 iters.add(msg.payload.makeIterator(command));
    -            sources[i] = msg.from;
    +
    +            Replica replica = replicaMap.get(msg.from);
    +            if (replica == null)
    --- End diff --
    
    It seems like we knew what kind of Replica it was when we sent the message. The kind of
Replica it might be when we get the response could be different? Probably not given how pending
states work.
    
    I just don't get how we could fail to decorate it. That's not right. If we sent the message
eliciting this response then we must have known at the time.


---

---------------------------------------------------------------------
To unsubscribe, e-mail: pr-unsubscribe@cassandra.apache.org
For additional commands, e-mail: pr-help@cassandra.apache.org


Mime
View raw message