dcapwell commented on code in PR #4202:
URL: https://github.com/apache/cassandra/pull/4202#discussion_r2193417436
##########
src/java/org/apache/cassandra/service/accord/serializers/TopologySerializers.java:
##########
@@ -159,4 +167,127 @@ public long serializedSize(Topology topology)
return size;
}
};
+
+ public static final UnversionedSerializer<Topology> compactTopology = new
UnversionedSerializer<>()
+ {
+ @Override
+ public void serialize(Topology topology, DataOutputPlus out) throws
IOException
+ {
+ out.writeLong(topology.epoch());
+ CollectionSerializers.serializeList(topology.staleIds(), out,
TopologySerializers.nodeId);
+
+ List<Shard> shards = topology.shards();
+
+ // need to loop twice; once to collect tables/ranges, and another
to save shards
+ ImmutableUniqueList<TableId> tables;
Review Comment:
> For deserialisation, the object->int mapping is unused, so this is a
costly waste of effort to construct.
Agree, on deserialize it could be `List` directly
> The normal approach here is to simply write them in the order they come,
and ensure they are inserted into the hash map in that same order, so that the
hash map contains their index. This saves an array allocation.
We then loose the header saying how many items exist, but you try to address
this in the last comment.
Ill take a look at option 1/2 and play around with this
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: pr-unsubscr...@cassandra.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: pr-unsubscr...@cassandra.apache.org
For additional commands, e-mail: pr-h...@cassandra.apache.org
