You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
yt-dlp/yt_dlp
Jody Bruchon a9ac178eb1 Use 64 KiB buffered writes for performance and less fragmentation
I use yt-dlp on Windows writing to a Linux system via SMB over a
10GbE connection and downloading via 400 Mbps cable internet. I
have observed that downloads often seem to start very fast (40+
MiB/sec) but then throttle down to 8-20 MiB/sec. I also observed
a large amount of disk thrashing for such a large array and small
amount of data that's supposedly being written sequentially.

The problem is two-fold. Downloaded fragments are stored using a
very short-lived *-FragX file, then immediately appended to the
stream upon fragment completion, and deleted. Both operations use
small write buffers. When the OS write buffers start to flush, the
two sets of writes plus the large number of writes start to force
competition to complete the queued writes in different areas of
the volume.

Python defaults to sending writes at the underlying device's
"block size" or a fallback to io.DEFAULT_BUFFER_SIZE. In practical
terms, this means a write buffer of 4096 or 8192 bytes. This
commit increases most write buffers to 65536 (64 KiB) using the
open() buffering=X option, significantly speeding up writes of
larger chunks of data and reducing potential fragmentation in low
disk space conditions. With these changes, I consistently see fast
downloads and the array thrashing is noticeably lessened.
4 months ago
..
__pyinstaller [rh:curlcffi] Add support for `curl_cffi` 1 year ago
compat [cleanup] Deprecate more compat functions (#11439) 5 months ago
dependencies [core] Catch broken Cryptodome installations (#11486) 5 months ago
downloader [cleanup] Bump ruff to 0.8.x (#11608) 4 months ago
extractor Use 64 KiB buffered writes for performance and less fragmentation 4 months ago
networking [cleanup] Misc (#10807) 7 months ago
postprocessor [cleanup] Bump ruff to 0.8.x (#11608) 4 months ago
utils Use 64 KiB buffered writes for performance and less fragmentation 4 months ago
YoutubeDL.py Use 64 KiB buffered writes for performance and less fragmentation 4 months ago
__init__.py [cleanup] Bump ruff to 0.8.x (#11608) 4 months ago
__main__.py [docs] Misc Cleanup (#8977) 1 year ago
aes.py [cleanup] Bump ruff to 0.8.x (#11608) 4 months ago
cache.py [cleanup] Add more ruff rules (#10149) 10 months ago
cookies.py [cookies] Add `--cookies-from-browser` support for MS Store Firefox (#11731) 4 months ago
jsinterp.py [jsinterp] Improve `slice` implementation (#10664) 8 months ago
minicurses.py
options.py [cleanup] Misc (#11554) 5 months ago
plugins.py [cleanup] Bump ruff to 0.8.x (#11608) 4 months ago
socks.py [cleanup] Add more ruff rules (#10149) 10 months ago
update.py [update] Check 64-bitness when upgrading ARM builds (#11819) 4 months ago
version.py Release 2024.12.13 4 months ago
webvtt.py [cleanup] Add more ruff rules (#10149) 10 months ago