xpra icon
Bug tracker and wiki

Opened 8 months ago

Closed 2 months ago

Last modified 2 months ago

#1926 closed enhancement (fixed)

replace numpy dependency in websockify

Reported by: Antoine Martin Owned by: J. Max Mena
Priority: major Milestone: 2.4
Component: external Version: 2.3.x
Keywords: Cc:

Description

As per ticket:1913#comment:2, importing numpy wastes a lot of memory and the only thing that websockify uses it for is to do a string XOR!

Let's submit a patch to make this more pluggable, we have our own code which can provide it. (our cyxor cython extension)
The import would need to be deferred since the websockify module imports the sub-modules. (ie: WebSocket constructor could get the xor wrapper function from websockify.xor which we would then be able to patch up)

Alternatively, we could use the modules context hack to temporarily hide "numpy" from the module loader, then patch the _unmask method.
This approach would work with older version of websockify - and they have not made any new releases in years now...

See also #1134.

Attachments (1)

cyxor-64-bit.patch (2.8 KB) - added by Antoine Martin 2 months ago.
64-bit is not faster!

Download all attachments as: .zip

Change History (9)

comment:1 Changed 8 months ago by Antoine Martin

Owner: changed from Antoine Martin to J. Max Mena

r19989 does this by pretending numpy is not installed and then monkeypatching a cython function to do the work.

This should be faster than the numpy code it replaces:

  • one fewer copy of the buffer data when doing the XOR
  • no longer copying the buffers when reading from the websocket (pass memoryview directly to caller, queue the rest)

The impact should be negligible since the server doesn't do a lot of reading.

importing websockify will now show this warning:

WARNING: no 'numpy' module, HyBi protocol will be slower

This can safely be ignored.

@maxmylyn: please check that this new code does not cause any performance regressions.

comment:2 Changed 7 months ago by Antoine Martin

r19989 caused a regression reported in ticket:1941#comment:7 :

disconnect: zlib packet decompression failed must be string or read-only buffer, not memoryview

Fixed in r20229.

comment:3 Changed 5 months ago by Antoine Martin

Another regression: #2001

comment:4 Changed 2 months ago by Antoine Martin

See also: #2104, with websockify > 0.8 we can silence the numpy warning.

comment:5 Changed 2 months ago by Antoine Martin

Resolution: fixed
Status: newclosed

Not heard back, so:

  • r21389 adds a simple command line tool to compare the performance of the default websockify unmask function (which uses numpy) and our cython implementation
  • r21392 adds a unit test to verify that we get the same result as numpy, no matter what the buffer alignment is

Sample output:

$ python ./tests/xpra/net/websockify_numpy_vs_cython.py 
* slice size: 1KB
 - websockify via numpy:      119MB/s
 - xpra cython         :      635MB/s
* slice size: 8KB
 - websockify via numpy:     1189MB/s
 - xpra cython         :     4016MB/s
* slice size: 64KB
 - websockify via numpy:     3148MB/s
 - xpra cython         :     9698MB/s
* slice size: 1024KB
 - websockify via numpy:     1207MB/s
 - xpra cython         :     6416MB/s
* slice size: 16384KB
 - websockify via numpy:     1096MB/s
 - xpra cython         :     1170MB/s
* slice size: 65536KB
 - websockify via numpy:     1152MB/s
 - xpra cython         :     1254MB/s

It's much faster on smaller packets (1KB: ~5 times faster!) than on really big ones (64MB: only 10%).
Typical packets are in the 1-64KB range and this is also where we benefit the most.


But I wasn't happy with those values as lz4 can compress data at that speed.
So r21393 optimizes the code for better performance.
Here are some results after setting cpu governor to performance:

$ python ./tests/xpra/net/websockify_numpy_vs_cython.py 
* slice size: 1KB
 - xpra cython         :      899MB/s
 - websockify via numpy:      141MB/s
* slice size: 8KB
 - xpra cython         :     5330MB/s
 - websockify via numpy:     1458MB/s
* slice size: 64KB
 - xpra cython         :    14233MB/s
 - websockify via numpy:     4138MB/s
* slice size: 1024KB
 - xpra cython         :    11877MB/s
 - websockify via numpy:     4809MB/s
* slice size: 16384KB
 - xpra cython         :     3974MB/s
 - websockify via numpy:     1173MB/s
* slice size: 65536KB
 - xpra cython         :     2833MB/s
 - websockify via numpy:     1166MB/s

Note: when numpy gets vectorized optimizations (https://lwn.net/Articles/691932/), we can revisit this.

Last edited 2 months ago by Antoine Martin (previous) (diff)

Changed 2 months ago by Antoine Martin

Attachment: cyxor-64-bit.patch added

64-bit is not faster!

comment:6 Changed 2 months ago by Antoine Martin

Minor enhancements in r21394 + r21395 + r21396. (unit tests, etc)

Not sure why, but the 64-bit version attached is not faster..

Last edited 2 months ago by Antoine Martin (previous) (diff)

comment:7 Changed 2 months ago by Antoine Martin

Fix for building on 32-bit windows with python3: r21454.

comment:8 Changed 2 months ago by Antoine Martin

r21465 adds support for websockify > 0.8 in the performance test (see #2104).
The results are the same or even lower than with websockify <= 0.8

See also #2121.

Last edited 2 months ago by Antoine Martin (previous) (diff)
Note: See TracTickets for help on using tickets.