[B! TCP][performance] y_uukiã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

y_uuki id:y_uuki

TCPã¨performanceã«é–¢ã™ã‚‹y_uukiã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (16)

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

${{author_name}}$
{{author_name}}{{created}}
{{ #comment }}{{ comment }}{{ /comment }}
- {{ label }}

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

TIME_WAITã«é–¢ã™ã‚‹è©± | GREE Engineering
ã“ã‚“ã«ã¡ã‚ã€‚ã›ã˜ã¾ã§ã™ã€‚ å…ˆæ—¥ãƒãƒã£ãŸ C302CA ãŒå±Šã„ãŸã‚“ã§ã™ãŒã€ï¼ˆè©±ãŒé•·ããªã‚‹ã®ã§ã™ã”ã„é›‘ã«ã„ã†ã¨ï¼‰Chromebook ã®ã€ã„ã‚ã„ã‚å‰ŠãŽè½ã¨ã—ã¦æœ€é©åŒ–ã™ã‚‹ã¨ã„ã†æ–¹å‘æ€§ã«ã¯ã€ã¨ã¦ã‚‚å¦ã¶ã¨ã“ã‚ãŒã‚ã‚Šã¾ã—ãŸã€‚ æ˜¨å¹´æœ«ã€ MySQLã®ã‚µãƒ¼ãƒé›†ç´„ã«å–ã‚Šçµ„ã‚“ã§ã„ã‚‹ã¨ã„ã†è©±ã‚’ã•ã›ã¦ã„ãŸã ãã¾ã—ãŸãŒã€ã€ŒDBé›†ç´„ã§ãã¦ããŸã®ã§ã€Webã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã‚µãƒ¼ãƒã‚‚é›†ç´„ã™ã‚‹ã‹ãƒ¼ã€ã¨ã„ã†ã“ã¨ã§ã€æœ€è¿‘ã¯Webã‚µãƒ¼ãƒã®é›†ç´„ã«ã‚‚å–ã‚Šçµ„ã‚“ã§ã„ã¾ã™ã€‚ ãŸã ã€PHPã‚´ãƒªã‚´ãƒªæ›¸ã„ã¦ã‚‹ã®ã§ã¯ãªãã€é›†ç´„ã—ã¦ã„ãä¸Šã§ Linux ã‚„ TCP çš„ãªè¦³ç‚¹ã‹ã‚‰ãƒãƒ¥ãƒ¼ãƒ‹ãƒ³ã‚°ã—ãŸã»ã†ãŒè‰¯ã„ã¨ã“ã‚ã‚’è¦‹ã¦ã¾ã—ã¦ã€ãã®ã¸ã‚“ã«ã¤ã„ã¦ã‚ã‚‹ã¦ã„ã©ã¾ã¨ã¾ã£ãŸã®ã§ã€æ¬¡ã®ã‚¹ãƒ©ã‚¤ãƒ‰ã‚’æ›¸ã‹ã›ã¦ã„ãŸã ãã¾ã—ãŸã€‚
y_uuki 2017/06/02
performance

TCP
ãƒªãƒ³ã‚¯
Developing the fastest HTTP/2 server
Presentation material for TokyoRubyKaigi11. Describes techniques used by H2O, including: techniques to optimize TCP for responsiveness, server-push and cache digests.Read less
y_uuki 2016/05/28
TCP send bufferã®æœ€é©åŒ–ãŠã‚‚ã—ã‚ã„

tcp

http2

performance
ãƒªãƒ³ã‚¯
mTCP: a Highly Scalable User-level TCP Stack for Multicore Systems | USENIX
EunYoung Jeong, Shinae Woo, Muhammad Jamshed, and Haewon Jeong, Korea Advanced Institute of Science and Techno logy (KAIST); Sunghwan Ihm, Princeton University; Dongsu Han and KyoungSoo Park, Korea Advanced Institute of Science and Techno logy (KAIST) Community Award Scaling the performance of short TCP connections on multicore systems is fundamentally challenging. Although many proposals have attem
y_uuki 2016/05/09
paper

tcp

performance
ãƒªãƒ³ã‚¯
TCP small queues [LWN.net]
The "bufferbloat" probl em is the result of excessive buffering in the network stack; it leads to long latencies and poor reliability in the network as a whole. Fixing it is a matter of buffering less data in each system between any two endpointsâ€”a task that sounds simple, but proves to be more challenging than one might expect. It turns out that buffering can show up in many surprising places in t
y_uuki 2016/03/30
ã‚ã¨ã§èªã‚€

linux

kernel

tcp

performance
ãƒªãƒ³ã‚¯
PVTCP: Towards Practical and Effective Congestion Control in Virtualized Datacenters
y_uuki 2016/03/30
xen

performance

tcp
ãƒªãƒ³ã‚¯
TCPã‚’ï¼ˆå°‘ã—ã¯ï¼‰ç†è§£ã—ã¦ãŠãã¹ããã®ç†ç”± | POSTD
ã“ã®è¨˜äº‹ã¯TCPã® å…¨ã¦ ã‚’ç†è§£ã™ã‚‹ã€ã‚ã‚‹ã„ã¯ ã€ŽTCP/IP Illustratedã€ ï¼ˆè¨³æ³¨ï¼šæ—¥æœ¬èªžç‰ˆ: ã€Žè©³è§£TCP/IPã€ˆVol.1ã€‰ãƒ—ãƒãƒˆã‚³ãƒ«ã€ ï¼‰ã‚’èªç ´ã—ã‚ˆã†ã¨ã‹ã€ãã†ã„ã†ã“ã¨ã§ã¯ã‚ã‚Šã¾ã›ã‚“ã€‚ã»ã‚“ã®å°‘ã—ã®TCPã®çŸ¥è˜ãŒã©ã‚Œã»ã©æ¬ ã‹ã›ãªã„ã‚‚ã®ãªã®ã‹ã«ã¤ã„ã¦ãŠè©±ã—ã¾ã™ã€‚ã¾ãšã¯ãã®ç†ç”±ã‚’ãŠè©±ã—ã¾ã—ã‚‡ã†ã€‚ ç§ãŒ Recurse Center ã§åƒã„ã¦ã„ã‚‹ã¨ãã€Pythonã§TCPã‚¹ã‚¿ãƒƒã‚¯ã‚’æ›¸ãã¾ã—ãŸï¼ˆ ã¾ãŸPythonã§TCPã‚¹ã‚¿ãƒƒã‚¯ã‚’æ›¸ã„ãŸã‚‰ã©ã†ãªã‚‹ã‹ã«ã¤ã„ã¦ã‚‚æ›¸ãã¾ã—ãŸ ï¼‰ã€‚ãã‚Œã¯ã¨ã¦ã‚‚æ¥½ã—ãã€ãŸã‚ã«ãªã‚‹çµŒé¨“ã§ã—ãŸã€‚ã¾ãŸãã‚Œã§ã„ã„ã¨æ€ã£ã¦ã„ãŸã‚“ã§ã™ã€‚ ãã“ã‹ã‚‰1å¹´ãã‚‰ã„çµŒã£ã¦ã€ä»•äº‹ã§ã€èª°ã‹ãŒã€ŒNSQã¸ãƒ¡ãƒƒã‚»ãƒ¼ã‚¸ã‚’é€ã£ãŸã‚“ã ãŒã€æ¯Žå›ž40ãƒŸãƒªç§’ã‹ã‹ã‚‹ã€ã¨Slackã«æŠ•ç¨¿ã—ã¦ã„ã‚‹ã®ã‚’è¦‹ã¤ã‘ã¾ã—ãŸã€‚ç§ã¯ã“ã®å•é¡Œã«ã¤ã„ã¦ã™ã§ã«1é€±é–“ã»ã©è€ƒãˆè¾¼ã‚“ã§ã„ã¾ã—ãŸãŒã€ã•ã£ã±ã‚Šç”ãˆãŒã§ã¾ã›ã‚“ã§ã—ãŸã€‚ ã“
y_uuki 2015/12/30
é…å»¶ACKã¨Nagleã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã®ç›¸æ€§å•é¡Œã€çŸ¥ã£ã¦ã¦ã‚‚ç¾å ´ã§åˆã‚ã¦è¦‹ã¦æ°—ã¥ãã®ã¯é›£ã—ãã†

TCP

performance
ãƒªãƒ³ã‚¯
User-space Network Processing
3. â€¢ B3 k6 z 2 l â€¢ mTCP +4Intel4DPDK wi â€¢ github mTCP+4DPDK orz â€¢ Key4Value4Store w k Linux l â€¢ RADIS â†’ â€¢ d v orz â€¢ Memcached â†’ â€¢ d 3 4. A G 7 LNPMXXT 4 0 0.2 0.4 0.6 0.8 1 0 20 40 60 80 100 Key size (bytes) Key size CDF by appearance USR APP ETC VAR SYS 0 0.2 0.4 0.6 0.8 1 1 10 100 1000 10000 100000 1e+06 Value size (bytes) Value Size CDF by appearance USR APP ETC VAR SYS 0 0.2 0.4 0.6 0.8 1 1 10
y_uuki 2015/12/19
èˆˆå‘³æ·±ã„

performance

TCP

memcached

kernel
ãƒªãƒ³ã‚¯
tcp/dccp: lockless listener [LWN.net]
netdev <net [email protected]>, Eric Dumazet <edumazet@google.com>, Ying Cai <ycai@google.com>, Willem de Bruijn <willemb@google.com>, Eric Dumazet <eric.dumazet@gmail.com> TCP listener refactoring : this is becoming interesting ! This patch series takes the steps to use normal TCP/DCCP ehash table to store SYN_RECV requests, instead of the private per-listener hash table we had until now. SYNACK
y_uuki 2015/11/11
TCP

kernel

performance
ãƒªãƒ³ã‚¯
Tuning TCP and NGINX on EC2
Chartbeat measures and monetizes attention on the web. They were experiencing slow load times and TCP retransmissions due to default system settings. Tuning various TCP, NGINX and EC2 ELB settings like increasing buffers, disabling Nagle's algorithm, and enabling HTTP keep-alive resolved the issues and improved performance. These included tuning settings like net.ipv4.tcp_max_syn_backlog, net.core
y_uuki 2015/09/26
performance

nginx

tcp
ãƒªãƒ³ã‚¯
Mobile TCP optimization - lessons learned in production
I did a keynote presentation at the SIGCOMM'15 HotMiddlebox workshop, "Mobile TCP optimization - Lessons Learned in Production". The title was set before I had any idea of what I'd really be talking about, just that it'd be about some of the stuff we've been working on at Teclo. So apologies if the content isn't an exact match for the title. This post contains my slides, interleaved with my speake
y_uuki 2015/08/26
tcp

performance
ãƒªãƒ³ã‚¯
Nginx Optimization: understanding sendfile, tcp_nodelay and tcp_nopush
This article a is translation by popular request of Optimisations Nginx, bien comprendre sendfile, tcpnodelay et tcpnopush I wrote in French in January. Most articles dealing with optimizing Nginx performances recommend to use sendfile, tcp_nodelay and tcp_nopush options in the nginx.conf configuration file. Unfortunately, almost none of them tell neither how they impact the Web server nor how the
y_uuki 2015/02/14
nginx

TCP

performance
ãƒªãƒ³ã‚¯
GitHub - satori-com/tcpkali: Fast multi-core TCP and WebSockets load generator.
You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session. You switched accounts on another tab or window. Reload to refresh your session. Dismiss alert
y_uuki 2015/02/11
performance

tcp
ãƒªãƒ³ã‚¯
wrkã§unix domain socketãªHTTPã‚µãƒ¼ãƒã‚’ãƒ™ãƒ³ãƒãƒžãƒ¼ã‚¯ - blog.nomadscafe.jp
wrkã«ç„¡ç†çŸ¢ç†ãªpatchã‚’ã‚ã¦ã¦ã€unix domain socketçµŒç”±ã§HTTPã‚µãƒ¼ãƒã‚’ãƒ™ãƒ³ãƒãƒžãƒ¼ã‚¯ã§ãã‚‹ã‚ˆã†ã«ã—ã¦ã¿ã¾ã—ãŸã€‚ kazeburo/wrk at unixdomain pullreqã¯ã—ã¦ãªã„ã€‚ Gazelleã‚„Rhebokã¨ã„ã£ãŸã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã‚µãƒ¼ãƒã‚’ä½œã£ã¦ã„ã¾ã™ãŒã€TCPçµŒç”±ã®ãƒ™ãƒ³ãƒãƒžãƒ¼ã‚¯ã§ã¯Ephemeral Portã®æž¯æ¸‡ã‚„TIME WAITã®ä¸Šé™ã«ã‚ãŸã£ã¦ã—ã¾ã„ã€ãƒ™ãƒ³ãƒãƒžãƒ¼ã‚¯ãŒã—ã¥ã‚‰ã„ã¨ã„ã†å•é¡ŒãŒã‚ã‚Šã¾ã™ã€‚ ãã“ã§nginxã‚’reverse proxyã¨ã—ã¦è¨ç½®ã—ã€nginxã¨ã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã‚µãƒ¼ãƒé–“ã‚’unix domain socketã§ç¹‹ã„ã§ãƒ™ãƒ³ãƒãƒžãƒ¼ã‚¯ã‚’ã¨ã£ã¦ã„ã¾ã—ãŸãŒã€nginxãŒãƒœãƒˆãƒ«ãƒãƒƒã‚¯ã«ãªã‚Šã‚„ã™ãã€ç›´æŽ¥ã‚¢ã‚¯ã‚»ã‚¹ã—ãŸã„ãªã¨è€ƒãˆã¦ã„ãŸã®ã§ã€ã‚„ã£ã¦ã¿ã¾ã—ãŸã€‚ ã“ã‚Œã‚’ä½¿ã£ã¦Rhebokã¨Unicornã® â€œHello Worldâ€ ãƒ™ãƒ³ãƒãƒžãƒ¼ã‚¯ã‚’è¡Œ
y_uuki 2015/01/28
performance

TCP
ãƒªãƒ³ã‚¯
mTCPä½¿ã£ã¦ã¿ãŸ
mTCP enables high-performance userspace TCP/IP stacks by bypassing the kernel and reducing system call overhead. It was shown to achieve up to 25x higher throughput than Linux for short flows. The document discusses porting the iperf benchmark to use mTCP, which required only minor changes. Performance tests found that mTCP-ified iperf achieved similar throughput as Linux iperf for different packe
y_uuki 2014/12/31
ã‚¹ãƒ«ãƒ¼ãƒ—ãƒƒãƒˆãŒãƒ‘ã‚±ãƒƒãƒˆã‚µã‚¤ã‚ºã«ä¾ã‚‰ãªã„ã®ã¯èˆˆå‘³æ·±ã„

performance

linux

kernel

tcp
ãƒªãƒ³ã‚¯
Xen-netback and xen-netfront multi-queue performance testing - Xen
The information in this page is out-dated. It doesn't reflect the latest performance of Xen network drivers. Please have a look at https://events.linuxfoundation.org/sites/events/files/slides/zoltan_kiss_xensummit2014_network_perf.pdf for latest information. Overview This page discusses the performance of an implementation of multiple transmit and receive queues (multiple shared rings) for xen-net
y_uuki 2014/05/12
xen

performance

tcp
ãƒªãƒ³ã‚¯
Engine Yard Blog
We are proud to be working with our customers while they continue to enjoy our Engine Yard products. Here are some changes we are making
y_uuki 2012/12/20
linux

tcp

network

performance
ãƒªãƒ³ã‚¯
1