Github user aweisberg commented on a diff in the pull request:

    https://github.com/apache/cassandra/pull/111#discussion_r120443366
  
    --- Diff: src/java/org/apache/cassandra/net/async/NettyFactory.java ---
    @@ -125,10 +126,19 @@
          * Determine the number of accept threads we need, which is based upon 
the number of listening sockets we will have.
          * We'll have either 1 or 2 listen sockets, depending on if we use SSL 
or not in combination with non-SSL. If we have both,
          * we'll have two sockets, and thus need two threads; else one socket 
and one thread.
    +     *
    +     * If the operator has configured multiple IP addresses (both {@link 
org.apache.cassandra.config.Config#broadcast_address}
    +     * and {@link org.apache.cassandra.config.Config#listen_address} are 
configured), then we listen on another set of sockets
    +     * - basically doubling the count. See CASSANDRA-9748 for more details.
          */
         static int determineAcceptGroupSize(InternodeEncryption 
internode_encryption)
         {
    -        return internode_encryption == InternodeEncryption.dc || 
internode_encryption == InternodeEncryption.rack ? 2 : 1;
    +        int listenSocketCount = internode_encryption == 
InternodeEncryption.dc || internode_encryption == InternodeEncryption.rack ? 2 
: 1;
    +
    +        if (MessagingService.shouldListenOnBroadcastAddress())
    +            listenSocketCount *= 2;
    --- End diff --
    
    Why do you need more threads to listen on more sockets? It's a non-blocking 
operation and more sockets doesn't indicate a need for more throughput?


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to