Charles-François Natali added the comment:
Alright, here's a simple patch bumping the buffers to 64K for asyncore
and asynchat.
> There is another problem with dispatcher_with_send: the buffering algorithm
> (both when appending and popping) is quadratic. You can easily observe it
> with your test script, when growing the DATA. async_chat looks much saner in
> that respect, I wonder why the same algorithm couldn't it be re-used.
Yeah, I noticed that.
But even in asynchat, there's a lot of copying going on, length
computations performed twice in a row, etc.
> (regardless, reading the asyncore code really hurts the eyes :-/)
Indeed.
----------
Added file: http://bugs.python.org/file28468/asyncore_buffsize.diff
_______________________________________
Python tracker <rep...@bugs.python.org>
<http://bugs.python.org/issue16787>
_______________________________________
diff --git a/Lib/asynchat.py b/Lib/asynchat.py
--- a/Lib/asynchat.py
+++ b/Lib/asynchat.py
@@ -56,8 +56,8 @@
# these are overridable defaults
- ac_in_buffer_size = 4096
- ac_out_buffer_size = 4096
+ ac_in_buffer_size = 65536
+ ac_out_buffer_size = 65536
# we don't want to enable the use of encoding by default, because that is a
# sign of an application bug that we don't want to pass silently
diff --git a/Lib/asyncore.py b/Lib/asyncore.py
--- a/Lib/asyncore.py
+++ b/Lib/asyncore.py
@@ -532,7 +532,7 @@
def initiate_send(self):
num_sent = 0
- num_sent = dispatcher.send(self, self.out_buffer[:512])
+ num_sent = dispatcher.send(self, self.out_buffer[:65536])
self.out_buffer = self.out_buffer[num_sent:]
def handle_write(self):
_______________________________________________
Python-bugs-list mailing list
Unsubscribe:
http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
