cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CASSANDRA-14556) Optimize streaming path in Cassandra
Date Fri, 13 Jul 2018 22:38:00 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-14556?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16543827#comment-16543827
] 

ASF GitHub Bot commented on CASSANDRA-14556:
--------------------------------------------

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

    https://github.com/apache/cassandra/pull/239#discussion_r202488581
  
    --- Diff: src/java/org/apache/cassandra/net/async/RebufferingByteBufDataInputPlus.java
---
    @@ -249,4 +250,42 @@ public ByteBufAllocator getAllocator()
         {
             return channelConfig.getAllocator();
         }
    +
    +    /**
    +     * Consumes bytes in the stream until the given length
    +     *
    +     * @param writer
    +     * @param len
    +     * @return
    +     * @throws IOException
    +     */
    +    public long consumeUntil(BufferedDataOutputStreamPlus writer, long len) throws IOException
    +    {
    +        long copied = 0; // number of bytes copied
    +        while (copied < len)
    +        {
    +            if (buffer.remaining() == 0)
    +            {
    +                try
    +                {
    +                    reBuffer();
    +                }
    +                catch (EOFException e)
    +                {
    +                    throw new EOFException("EOF after " + copied + " bytes out of " +
len);
    +                }
    +                if (buffer.remaining() == 0)
    +                    return copied == 0 ? -1 : copied;
    --- End diff --
    
    Here - https://github.com/dineshjoshi/cassandra/blob/f9573ccef63535b787e91e46fa886f3e3ff2eee8/src/java/org/apache/cassandra/io/sstable/format/big/BigTableBlockWriter.java#L213
    
    `reBuffer()` could likely fail due to network or channel being closed by the other side.
It could be treated as a normal `IOException` as we're not guaranteed that we would get those
many bytes from the other end. It feels a bit weird that rest of `RebufferingByteBufDataInputPlus`
throws `EOFException` and this method would throw an `AssertionError` when there may be a
network issue.



> Optimize streaming path in Cassandra
> ------------------------------------
>
>                 Key: CASSANDRA-14556
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-14556
>             Project: Cassandra
>          Issue Type: Improvement
>          Components: Streaming and Messaging
>            Reporter: Dinesh Joshi
>            Assignee: Dinesh Joshi
>            Priority: Major
>              Labels: Performance
>             Fix For: 4.x
>
>
> During streaming, Cassandra reifies the sstables into objects. This creates unnecessary
garbage and slows down the whole streaming process as some sstables can be transferred as
a whole file rather than individual partitions. The objective of the ticket is to detect when
a whole sstable can be transferred and skip the object reification. We can also use a zero-copy
path to avoid bringing data into user-space on both sending and receiving side.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

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


Mime
View raw message