Skip to content

Commit

Permalink
Add (larger) benchmarks with the rapidgzip zlib-support branch
Browse files Browse the repository at this point in the history
  • Loading branch information
@mxmlnkn
mxmlnkn committed Dec 27, 2023
1 parent 72b51e3 commit c59b353
Show file tree
Hide file tree
Showing 2 changed files with 109 additions and 34 deletions.
20 changes: 20 additions & 0 deletions README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -34,6 +34,26 @@ print(zlibng_concat_decode(c0 + c1)) # returns b"abcefg"

### benchmarks

To reproduce:

```bash
git clone --recursive https://github.com/zlib-ng/zlib-ng.git
cd zlib-ng/ && mkdir build && cd build
cmake -DCMAKE_INSTALL_PREFIX=/opt/zlib-ng ..
make -j $( nproc ) install
export CPATH=/opt/zlib-ng/include
export LD_LIBRARY_PATH="/opt/zlib-ng/lib:$LD_LIBRARY_PATH"
export LIBRARY_PATH="/opt/zlib-ng/lib:$LIBRARY_PATH"

python3 -m pip install --force-reinstall 'git+https://github.com/mxmlnkn/indexed_bzip2.git@zlib-support#egginfo=rapidgzip&subdirectory=python/rapidgzip'

git clone --recursive https://github.com/ap--/concatenated-zlib.git
cd concatenated-zlib/
python3 -m pip install --user .
cd bench
python time_load_chunks.py
```

To reproduce:
```shell
# ON A MAC YOU CAN INSTALL THE DEPENDECIES VIA homebrew
Expand Down
123 changes: 89 additions & 34 deletions bench/time_load_chunks.py
View file
Original file line number Diff line number Diff line change
@@ -1,9 +1,12 @@
import contextlib
import io
import itertools
import os
import time
import zlib

import numpy as np
import rapidgzip

from concatenated_zlib import zlib_concat_decode
from concatenated_zlib import zlibng_concat_decode
Expand Down Expand Up @@ -40,51 +43,103 @@ def timeit(label, div=1):
try:
yield
finally:
print(label, "took", (time.monotonic() - t0) / div, "seconds")
print(label, "took", round((time.monotonic() - t0) / div, 3), "seconds")


def bench():

c0 = open("./chunk_0000.dat", "rb").read()
c1 = open("./chunk_0001.dat", "rb").read()

with timeit("overhead", div=100):
for _ in range(100):
x = b"".join(itertools.islice(itertools.cycle([c0, c1]), 1000))

with timeit("stdlib.zlib", div=100):
for _ in range(100):
_data0 = b"".join(
zlib.decompress(c)
for c in itertools.islice(itertools.cycle([c0, c1]), 1000)
)

with timeit("zlib_concat_decode", div=100):
for _ in range(100):
_data1 = zlib_concat_decode(b"".join(itertools.islice(itertools.cycle([c0, c1]), 1000)))

with timeit("zlibng_concat_decode", div=100):
for _ in range(100):
_data2 = zlibng_concat_decode(b"".join(itertools.islice(itertools.cycle([c0, c1]), 1000)))

with timeit("zlib_multi_decompress", div=10):
for _ in range(10):
_data3 = zlib_multi_decompress(itertools.islice(itertools.cycle([c0, c1]), 1000))

with timeit("libdeflate_zlib_decompress", div=100):
test_bytes_list = lambda: itertools.islice(itertools.cycle([c0, c1]), 40000)
test_bytes_concatenated = b"".join(test_bytes_list())
print(f"Concatenated size: {len(test_bytes_concatenated) / 1e6} MB")

outputFilePath = "many-concatenated.zlib"
with open(outputFilePath, 'wb') as file:
file.write(test_bytes_concatenated)

with timeit("overhead"):
x = test_bytes_concatenated

with timeit("stdlib.zlib"):
_data0 = b"".join(zlib.decompress(c) for c in test_bytes_list())

with timeit("stdlib.zlib without joining"):
for c in test_bytes_list():
zlib.decompress(c)

for parallelization in [1, 2, 4, 8, os.cpu_count()]:
chunkSize = 1024 * 1024
with timeit(f"rapidgzip_decompress from BytesIO read and discard {chunkSize / 1024 / 1024} MiB chunks "
f"parallelization={parallelization}"):
with rapidgzip.RapidgzipFile(io.BytesIO(test_bytes_concatenated), parallelization=parallelization) as file:
while result := file.read(chunkSize):
pass

for parallelization in [1, 2, 4, 8, os.cpu_count()]:
with timeit(f"rapidgzip_decompress from BytesIO parallelization={parallelization}"):
with rapidgzip.RapidgzipFile(io.BytesIO(test_bytes_concatenated), parallelization=parallelization) as file:
_data5 = file.read()
assert _data0 == _data5
del _data5

for parallelization in [1, 2, 4, 8, os.cpu_count()]:
with timeit(f"rapidgzip_decompress from file parallelization={parallelization}"):
with rapidgzip.RapidgzipFile(outputFilePath, parallelization=parallelization) as file:
_data5 = file.read()
assert _data0 == _data5
del _data5

with timeit("zlib_concat_decode"):
_data1 = zlib_concat_decode(test_bytes_concatenated)
assert _data0 == _data1
del _data1

with timeit("zlibng_concat_decode"):
_data2 = zlibng_concat_decode(test_bytes_concatenated)
assert _data0 == _data2
del _data2

with timeit("zlib_multi_decompress"):
_data3 = zlib_multi_decompress(test_bytes_list())
assert _data0 == _data3
del _data3

with timeit("libdeflate_zlib_decompress"):
out = np.empty((1024*12 + 24,), dtype=np.uint8)
for _ in range(100):
_data4 = b"".join(
libdeflate_zlib_decode(c, out)
for c in itertools.islice(itertools.cycle([c0, c1]), 1000)
)
_data4 = b"".join(libdeflate_zlib_decode(c, out) for c in test_bytes_list())
assert _data0 == _data4
del _data4

# with timeit("onecall"):
# z.zlib_decode(b"".join(itertools.islice(itertools.cycle([c0, c1]), 256000)))

assert _data0 == _data1 == _data2 == _data3 == _data4


if __name__ == "__main__":
test()
bench()
bench()


# Concatenated size: 196.92 MB
# overhead took 0.0 seconds
# stdlib.zlib took 1.97 seconds
# stdlib.zlib without joining took 1.422 seconds
# rapidgzip_decompress from BytesIO read and discard 1.0 MiB chunks parallelization=1 took 1.406 seconds
# rapidgzip_decompress from BytesIO read and discard 1.0 MiB chunks parallelization=2 took 0.775 seconds
# rapidgzip_decompress from BytesIO read and discard 1.0 MiB chunks parallelization=4 took 0.439 seconds
# rapidgzip_decompress from BytesIO read and discard 1.0 MiB chunks parallelization=8 took 0.365 seconds
# rapidgzip_decompress from BytesIO read and discard 1.0 MiB chunks parallelization=16 took 0.361 seconds
# rapidgzip_decompress from BytesIO parallelization=1 took 2.098 seconds
# rapidgzip_decompress from BytesIO parallelization=2 took 1.603 seconds
# rapidgzip_decompress from BytesIO parallelization=4 took 1.048 seconds
# rapidgzip_decompress from BytesIO parallelization=8 took 1.166 seconds
# rapidgzip_decompress from BytesIO parallelization=16 took 1.312 seconds
# rapidgzip_decompress from file parallelization=1 took 2.088 seconds
# rapidgzip_decompress from file parallelization=2 took 1.669 seconds
# rapidgzip_decompress from file parallelization=4 took 0.988 seconds
# rapidgzip_decompress from file parallelization=8 took 1.133 seconds
# rapidgzip_decompress from file parallelization=16 took 1.27 seconds
# zlib_concat_decode took 1.954 seconds
# zlibng_concat_decode took 1.635 seconds
# zlib_multi_decompress took 2.018 seconds
# libdeflate_zlib_decompress took 1.374 seconds

0 comments on commit c59b353

Please sign in to comment.