Comment #5 on issue 374 by [email protected]: Add method in ByteString to
avoid big buffer copy.
http://code.google.com/p/protobuf/issues/detail?id=374
We've used Benoit's approach in core Apache code. Our final version is:
package com.google.protobuf; // This is a lie.
public final class ZeroCopyLiteralByteString extends LiteralByteString {
private ZeroCopyLiteralByteString() { super(null);}
public static ByteString wrap(final byte[] array) {
return new LiteralByteString(array);
// return ByteString.copyFrom(array);
}
public static ByteString wrap(final byte[] array, int offset, int length)
{
return new BoundedByteString(array, offset, length);
// return ByteString.copyFrom(array, offset, length);
}
}
We we change this to use the official ByteString.copyFrom, using the
standard YCSB perf tool, we see a degradation of performance of 22%. That
is, just using the commented lines of code above cost 22% of messages
managed per second (from 167000 to 130000).
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings
--
You received this message because you are subscribed to the Google Groups "Protocol
Buffers" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/protobuf.
For more options, visit https://groups.google.com/groups/opt_out.
