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

y_uuki id:y_uuki

TCPã¨linuxã«é–¢ã™ã‚‹y_uukiã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (60)

${{author_name}}$

{{{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©±
3. è‡ªå·±ç´¹ä»‹ - ã¾ãã¾ã MySQL ã§ã”é£¯é£Ÿã¹ã¦ã¾ã™ - ä¸€æ™‚æœŸã¯ Resource Monitoring ã‚„ KVS ã«ã‚‚ åŠ›å…¥ã‚Œã¦ã¾ã—ãŸ - ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯çš„ã«ã¯ç´ äººã§ã™ - Linuxã¨ãƒãƒ¼ãƒ‰ã‚¦ã‚§ã‚¢ã¯å—œã‚€ç¨‹åº¦ - disk I/O ã«ã¯ã‚€ã‹ã—ã‹ã‚‰èˆˆå‘³ã‚ã‚Šã¾ã™ - ãã®ä»– slideshare ã¯ã“ã¡ã‚‰ - http://www.slideshare.net/takanorisejima/ 4. æœ¬æ—¥ã®ãŠé¡Œ - kernel æ–°ã—ãã—ãŸã‚Šã™ã‚‹ã¨ã€TCPçš„ã«æ„è˜ã—ãŸã» ã†ãŒè‰¯ã„å¤‰åŒ–ãŒè¦‹ã¤ã‹ã‚‹ã®ã§ - ä»Šæ—¥ã¯ã€Webã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã‚µãƒ¼ãƒã®è¦³ç‚¹ã‹ ã‚‰ã€ connect(2) ã™ã‚‹éš›ã«æ°—ã«ãªã‚‹ TIME_WAIT ã«ã¤ã„ã¦ã€æ›¸ã„ã¦ã¿ã‚ˆã†ã‹ã¨æ€ã„ã¾ã™ - æœ‰è˜è€…ã‹ã‚‰ã®ãƒžã‚µã‚«ãƒªã‚’ã€å¼·ãæ“è¿Žã„ãŸã—ã¾ã™ 5. æœ€åˆã«å‚è€ƒè³‡æ–™ - ã“ã®äºŒã¤ã®è¨˜äº‹ã‚’èªã‚“ã§ã„ãŸã ã‘ã°ã€ãã‚Œã§æ¦‚ ãè‰¯ã„ã¨æ€ã†ã‚“ã§ã™ãŒ
y_uuki 2017/10/23
kernel

TCP
ãƒªãƒ³ã‚¯
tcp_bbr: add BBR congestion control Â· torvalds/linux@0f8782e
y_uuki 2017/07/27
TCP

linux
ãƒªãƒ³ã‚¯
é«˜é€Ÿã«ãƒªãƒ¢ãƒ¼ãƒˆãƒ›ã‚¹ãƒˆã®ãƒãƒ¼ãƒˆãŒListenã—ã¦ã„ã‚‹ã‹ã‚’èª¿ã¹ã‚‹ - äººé–“ã¨ã‚¦ã‚§ãƒ–ã®æœªæ¥
hb.matsumoto-r.jp ä»¥ä¸‹ã®ã‚¨ãƒ³ãƒˆãƒªã¯ä¸€éƒ¨èª¤èªãŒå«ã¾ã‚Œã¦ã„ãŸã®ã§ã€ä¸Šè¨˜ã‚¨ãƒ³ãƒˆãƒªã«ãã®æ—¨ã‚’ã¾ã¨ã‚ã¾ã—ãŸã®ã§å¾¡è¦§ãã ã•ã„ã€‚ ã¨ã‚ã‚‹äº‹æƒ…ã§ãƒŸãƒ‰ãƒ«ã‚¦ã‚§ã‚¢ä¸Šã‹ã‚‰é«˜é€Ÿã«ãƒªãƒ¢ãƒ¼ãƒˆãƒ›ã‚¹ãƒˆã®ãƒãƒ¼ãƒˆã®Listenãƒã‚§ãƒƒã‚¯ã‚’ã—ãŸããªã‚Šã¾ã—ãŸã€‚ãƒãƒ¼ã‚«ãƒ«ãƒ›ã‚¹ãƒˆã®ãƒãƒ¼ãƒˆã§ã‚ã‚Œã°ã€/procã‚„netlinkãªã©ã‚’ä½¿ã£ã¦ç´ æ—©ããƒã‚§ãƒƒã‚¯ã™ã‚‹æ–¹æ³•ãŒã‚ã‚Šã¾ã™ãŒã€ä»Šå›žã¯å¯¾è±¡ãŒãƒªãƒ¢ãƒ¼ãƒˆãƒ›ã‚¹ãƒˆãªã®ã§ã‚½ã‚±ãƒƒãƒˆã§ãªã‚“ã¨ã‹ã™ã‚‹å¿…è¦ãŒã‚ã‚Šã¾ã™ã€‚ ãã“ã§ã€èª°ã‚‚ãŒã¾ãšæ€ã„ã¤ãã®ã¯ã€connect()ã‚·ã‚¹ãƒ†ãƒ ã‚³ãƒ¼ãƒ«ã«ã‚ˆã£ã¦ãƒªãƒ¢ãƒ¼ãƒˆãƒ›ã‚¹ãƒˆã®ãƒãƒ¼ãƒˆã«æŽ¥ç¶šã—ã«ã„ã£ã¦ã€connectã§ãã‚Œã°OKã€ã§ããªã‘ã‚Œã°NGã¨åˆ¤å®šã™ã‚‹æ–¹æ³•ãŒã‚ã‚Šå¾—ã‚‹ã§ã—ã‚‡ã†ã€‚ï¼ˆé«˜è² è·æ™‚ã«æŽ¥ç¶šã§ããªã„ãƒ‘ã‚¿ãƒ¼ãƒ³ã¯Listenã—ã¦ã„ãªã„ã¨åˆ¤å®šã—ã¦ã‚ˆã„ï¼‰ ãã“ã§ä¸€æ—¦ã€æœ€ä½Žé™socket()ã‚·ã‚¹ãƒ†ãƒ ã‚³ãƒ¼ãƒ«ã¨connect()ã‚·ã‚¹ãƒ†ãƒ ã‚³ãƒ¼ãƒ«ã§æŽ¥ç¶šã™ã‚‹æ™‚ã®ãƒ‘ã‚±ãƒƒãƒˆã‚’tcpdumpã§çœº
y_uuki 2017/02/13
ã‚«ãƒ¼ãƒãƒ«ãŒå‹æ‰‹ã«rstãƒ‘ã‚±ãƒƒãƒˆé€ã‚‹ã‚„ã¤ã€å¦ç”Ÿã®ã¨ãã«éé‡ã—ã¦æ³£ããã†ã«ãªã£ãŸ

Linux

network

TCP
ãƒªãƒ³ã‚¯
91ç²¾å“ä¸€åŒº_å›½äº§ä¸€çº§å¼º ç‰‡åœ¨çº¿è§‚çœ‹AV_äºšæ´²ç²¾å“æ— ç åˆå¤œç¦åˆ©ä¸æ–‡å—å¹•_ä¹…ä¹…ä¹…ä¹…ç²¾å“å›½äº§å…è´¹
2023å…æœˆæ–°æµå‡ºé»‘å®¢ç ´è§£å®¶åºç½‘ç»œæ‘„åƒå¤´ç›‘æŽ§å·æ‹??è€å¤«å«©å¦»æ—¥å¸¸æˆ¿äº‹æ€§ç”Ÿæ´»éªšåª³å¦‡ç»å¸¸å–œæ¬¢è¶´åœ¨ä¸Šé¢è‰è€å…¬
y_uuki 2016/04/29
Linux

TCP
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‚¯
TCP Keepalive HOWTO
This document describes the TCP keepalive implementation in the linux kernel, introduces the overall concept and points to both system configuration and software development. Table of Contents1. Introduction1.1. Copyright and License1.2. Disclaimer1.3. Credits / Contributors1.4. Feedback1.5. Translations2. TCP keepalive overview2.1. What is TCP keepalive?2.2. Why use TCP keepalive?2.3. Checking fo
y_uuki 2015/12/31
TCP

Linux

kernel
ãƒªãƒ³ã‚¯
redir(1): redirect tcp connections - Linux man page
redir(1) - Linux man page Name redir - redirect tcp connections Synopsis redir [--laddr=incoming.ip.address] [--caddr=host] [--debug] [--syslog [--name=str] [--timeout=n] [--bind_addr=my.other.ip.address] [--ftp=type] [--transproxy] [--connect=host:port] --lport=port --cport=port [--bufsize=n] [--max_bandwidth=n] [--random_wait=n] [--wait_in_out=n] redir --inetd [--caddr=host] [--debug] [--syslog]
y_uuki 2015/12/21
linux

tcp
ãƒªãƒ³ã‚¯
How TCP backlog works in Linux
When an application puts a socket into LISTEN state using the listen syscall, it needs to specify a backlog for that socket. The backlog is usually described as the limit for the queue of incoming connections. Because of the 3-way handshake used by TCP, an incoming connection goes through an intermediate state SYN RECEIVED before it reaches the ESTABLISHED state and can be returned by the accept s
y_uuki 2015/08/02
tcp

linux

kernel
ãƒªãƒ³ã‚¯
tcptrackã§ãƒˆãƒ©ãƒ•ã‚£ãƒƒã‚¯ã‚„é€šä¿¡çŠ¶æ³ã‚’ãƒªã‚¢ãƒ«ã‚¿ã‚¤ãƒ æŠŠæ¡ | Pocketstudio.jp log3
â—†tcptrackã¯ã€ã„ã‚ã‚†ã‚‹snifferã®ï¼‘ã¤ã€‚ ã€Œã‚‚ã—ã‹ã—ã¦ Dos æ”»æ’ƒã‚’é£Ÿã‚‰ã£ã¦ã‚‹ï¼Ÿã€ã¨ã„ã†æ™‚ã€ã¿ãªã•ã‚“ã©ã†ã•ã‚Œã¦ã„ã¾ã™ï¼Ÿ Linux ã‚µãƒ¼ãƒã® TCPÂ é€šä¿¡çŠ¶æ³ã‚’æŠŠæ¡ã™ã‚‹ã®ã«ã€æŽ¥ç¶šå…ƒãƒ»æŽ¥ç¶šå…ˆã®æŠŠæ¡ã§ã‚ã‚Œã°ã€netstat ã‚’ä½¿ã†ã®ãŒä¸€èˆ¬çš„ã€‚ã‚ˆã‚Šç´°ã‹ãªé€šä¿¡çŠ¶æ³ã§ã‚ã‚Œã° tcpdump ã‚’ä½¿ã†ã§ã—ã‚‡ã†ã€‚ã§ã™ãŒã€é€šä¿¡çŠ¶æ³ã ã‘ã§ãªãã€â€è»¢é€é‡â€ã‚’çŸ¥ã‚ŠãŸã„å ´åˆã‚„ã€ä»Šã©ã“ã®ãƒ›ã‚¹ãƒˆã¨ã®é€šä¿¡ãŒã‚µãƒ¼ãƒã®è² æ‹…ã«ãªã£ã¦ã„ã‚‹ã‹ã‚’æŠŠæ¡ã™ã‚‹äº‹ã¯ã€ãã†ãã†å˜ç´”ã§ã¯ã‚ã‚Šã¾ã›ã‚“ã€‚ ãã‚“ãªã¨ã tcptrack ã‚’ä½¿ã†ã®ãŒä¾¿åˆ©ã§ã™ã€‚tcptrack ã¯ã€â€topâ€ ã‚³ãƒžãƒ³ãƒ‰ã®ã‚ˆã†ã«ã€ç§’å˜ä½ã§åˆ»ã€…ã¨é€šä¿¡çŠ¶æ³ã‚’è¡¨ç¤ºã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ã©ã®ãƒ›ã‚¹ãƒˆé–“ã§ã€ã©ã®ã‚ˆã†ãªé€šä¿¡çŠ¶æ…‹ã§ã€ã‚¢ã‚¤ãƒ‰ãƒ«çŠ¶æ…‹ã®ç§’æ•°ã€è»¢é€ãƒ¬ãƒ¼ãƒˆã‚’è¡¨ç¤ºã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ â—†ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«æ–¹æ³• ãƒã‚¤ãƒŠãƒªã¾ãŸã¯ã‚½ãƒ¼ã‚¹ã‚³ãƒ¼ãƒ‰ã‹ã‚‰ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã§ãã¾ã™ã€‚ç¾è¡Œãƒãƒ¼ã‚¸ãƒ§ãƒ³
y_uuki 2015/07/18
linux

tcp
ãƒªãƒ³ã‚¯
TCP keepaliveè¨å®šã§ãƒãƒ¼ãƒ•ã‚³ãƒã‚¯ã‚·ãƒ§ãƒ³ã‚’è§£æ¶ˆ
ï¼’å°ã®ã‚µãƒ¼ãƒé–“ãŒTCPãƒãƒ¼ãƒ•ã‚³ãƒã‚¯ã‚·ãƒ§ãƒ³çŠ¶æ…‹ã«ãªã‚‹äº‹è±¡ãŒç™ºç”Ÿã—ãŸã€‚ TCP keepalive(ã‚ãƒ¼ãƒ—ã‚¢ãƒ©ã‚¤ãƒ–)ã®è¨å®šã§ãã®çŠ¶æ…‹ã‚’è§£æ¶ˆã™ã‚‹ãŸã‚ã®æ‰‹æ®µã‚’ãƒ¡ãƒ¢ã—ã¦ãŠãã€‚ â—† æ§‹æˆ ã‚µãƒ¼ãƒAã¨ã‚µãƒ¼ãƒB â—† ã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ä»•æ§˜ ã‚µãƒ¼ãƒAãŒã‚µãƒ¼ãƒBä¸Šã®ãƒ•ã‚¡ã‚¤ãƒ«ã‚’å®šæœŸçš„ã«å–å¾—ã™ã‚‹ã€‚ ä¸¡ã‚µãƒ¼ãƒé–“ã¯TCPã‚³ãƒã‚¯ã‚·ãƒ§ãƒ³ã‚’å¼µã‚Šç¶šã‘ã¦ã„ã‚‹ã€‚é€šä¿¡ã”ã¨ã«æŽ¥ç¶šã™ã‚‹ä»•æ§˜ã§ã¯ãªã„ã€‚ â—† ãƒˆãƒ©ãƒ–ãƒ« ã‚µãƒ¼ãƒBã«éšœå®³ãŒç™ºç”Ÿã—ã€ã‚µãƒ¼ãƒBãŒçªç„¶ãƒ€ã‚¦ãƒ³ã—ãŸã€‚ ãã®ãŸã‚ã‚µãƒ¼ãƒBã‹ã‚‰ã‚µãƒ¼ãƒAã¸ã®TCPã®ã‚¯ãƒãƒ¼ã‚ºå‡¦ç†ãŒè¡Œã‚ã‚Œãšã€ ã‚µãƒ¼ãƒAã ã‘ã«TCPã®çŠ¶æ…‹ãŒESTABLISHEDã§æ®‹ã‚‹ãƒãƒ¼ãƒ•ã‚³ãƒã‚¯ã‚·ãƒ§ãƒ³ã¨ãªã£ã¦ã—ã¾ã£ãŸã€‚ LANã‚±ãƒ¼ãƒ–ãƒ«ãŒæŠœã‘ãŸå ´åˆãªã©ã‚‚åŒã˜çŠ¶æ…‹ã«ãªã‚‹ã ã‚ã†ã‹ã€‚ ãƒãƒ¼ãƒ•ã‚³ãƒã‚¯ã‚·ãƒ§ãƒ³ã®æœ‰ç„¡ã¯ã€ã‚µãƒ¼ãƒAã€ã‚µãƒ¼ãƒBã§ãã‚Œãžã‚Œ "lsof -n"ã€"netstat -an" ãªã©ã®ã‚³ãƒžãƒ³ãƒ‰ã‚’æ‰“ã¡ã€çŠ¶æ…‹ãŒESTABLISHEDã«ãª
y_uuki 2015/07/02
TCP

kernel

linux
ãƒªãƒ³ã‚¯
conntrack ãŒãƒãƒ¼ãƒ‰ã•ã‚ŒãŸã‚µãƒ¼ãƒã§ LVS-DR ã‚’æ§‹æˆã™ã‚‹å ´åˆã®æ³¨æ„ç‚¹ - Cybozu Inside Out | ã‚µã‚¤ãƒœã‚¦ã‚ºã‚¨ãƒ³ã‚¸ãƒ‹ã‚¢ã®ãƒ–ãƒã‚°
ã“ã‚“ã«ã¡ã¯ã€‚ã‚¤ãƒ³ãƒ•ãƒ©é–‹ç™ºãƒãƒ¼ãƒ ã®æ·±è°·ã§ã™ã€‚ Linux ã® conntrack ãƒ¢ã‚¸ãƒ¥ãƒ¼ãƒ«ãŒãƒãƒ¼ãƒ‰ã•ã‚ŒãŸã‚µãƒ¼ãƒãƒ¼ã§ã€ãƒ¬ã‚¤ãƒ¤ãƒ¼4 (L4) ãƒãƒ¼ãƒ‰ãƒãƒ©ãƒ³ã‚µã‚’ LVS-DR ã§æ§‹æˆã—ãŸã¨ã“ã‚ã€ ãƒ‘ã‚±ãƒƒãƒˆãƒ•ã‚£ãƒ«ã‚¿ãŒæ£å¸¸ã«åƒã‹ãšã€é€šä¿¡ã¯ã§ãã‚‹ã‚‚ã®ã®ã€ä¸€éƒ¨ã®ãƒ‘ã‚±ãƒƒãƒˆã¯è½ã¡ã‚‹ã¨ã„ã†æŒ™å‹•ã«æ‚©ã¾ã•ã‚Œã¾ã—ãŸã€‚ ä»Šå›žã¯ã“ã®å•é¡Œã«ã¤ã„ã¦è©³ç´°ã«è§£æžã—ã€åŽŸå› ã¨å¯¾ç–ãŒã‚ã‹ã£ãŸã®ã§ã”ç´¹ä»‹ã„ãŸã—ã¾ã™ã€‚ ç™ºç”Ÿã—ãŸå•é¡Œ ä»Šå›žã®å•é¡Œã¯ã€Ubuntu ã‚’ä½¿ã£ã¦ L4 ãƒãƒ¼ãƒ‰ãƒãƒ©ãƒ³ã‚µã‚’å®Ÿè£…ã—ã¦ã„ã‚‹æ™‚ã«éé‡ã—ã¾ã—ãŸã€‚ æ§‹ç¯‰ã—ãŸãƒãƒ¼ãƒ‰ãƒãƒ©ãƒ³ã‚µã«ã¯ã€ã‚»ã‚ãƒ¥ãƒªãƒ†ã‚£å¯¾ç–ã®ãŸã‚ iptables ã®ãƒ©ãƒƒãƒ‘ãƒ¼ã§ã‚ã‚‹ UFW ã‚’ä½¿ã£ã¦ã€ ãƒ‘ã‚±ãƒƒãƒˆãƒ•ã‚£ãƒ«ã‚¿ã‚’ã‹ã‘ã¦ã„ã¾ã—ãŸã€‚ ãƒãƒ¼ãƒ‰ãƒãƒ©ãƒ³ã‚µã®å®Ÿè£…ã«ã¯ã€Linux ã‚«ãƒ¼ãƒãƒ«ã«å®Ÿè£…ã•ã‚Œã¦ã„ã‚‹ LVS ã‚’ä½¿ã„ã¾ã—ãŸã€‚ LVS ã§ã¯ã„ãã¤ã‹ã®ãƒ‘ã‚±ãƒƒãƒˆè»¢é€æ–¹å¼ã‚’ã‚µãƒãƒ¼ãƒˆã—ã¦ãŠã‚Šã€ä»Šå›žã¯ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹ã«å„ªã‚Œã‚‹
y_uuki 2015/06/14
lvs

TCP

kernel

linux
ãƒªãƒ³ã‚¯
The SO_REUSEPORT socket option [LWN.net]
One of the features merged in the 3.9 development cycle was TCP and UDP support for the SO_REUSEPORT socket option; that support was implemented in a series of patches by Tom Herbert. The new socket option allows multiple sockets on the same host to bind to the same port, and is intended to improve the performance of multithreaded network server applications running on top of multicore systems. Th
y_uuki 2015/05/16
linux

kernel

tcp
ãƒªãƒ³ã‚¯
hiboma/kernel/net/net-backlog.md at master Â· hiboma/hiboma
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 2015/04/27
linux

kernel

TCP
ãƒªãƒ³ã‚¯
Linuxã«ãŠã‘ã‚‹syncookiesã®å®Ÿè£… - yuya_prestoã®å¤ã„æ—¥è¨˜ï¼ˆã¯ã¦ãªãƒ€ã‚¤ã‚¢ãƒªãƒ¼ã‹ã‚‰ç§»è¡Œï¼‰
ã‚ã¾ã‚Šã«ã‚‚å¿™ã—ã„ã®ã§ã€ã¨ã‚Šã‚ãˆãšã€‚ã“ã®è¨˜äº‹ã¯å¾Œã§ä¿®æ£ã™ã‚‹ï¼ ã‚µãƒ¼ãƒã«å¤§é‡ã®SYNã‚’é€ã‚Šã¤ã‘ã‚‹ã“ã¨ã§ã€ã‚µãƒ¼ãƒå´ã§æŽ¥ç¶šå¾…ã¡çŠ¶æ…‹ã®ã‚³ãƒã‚¯ã‚·ãƒ§ãƒ³ã‚’ä¿æŒã™ã‚‹ã‚ãƒ¥ãƒ¼ã‚’ã‚ãµã‚Œè¿”ã•ã›ã‚‹ï¼ˆã¤ã¾ã‚Šãã‚Œä»¥é™ã®æŽ¥ç¶šè¦æ±‚ã¯ç ´æ£„ã•ã‚Œã‚‹ï¼‰ã“ã¨ã§ã€æŽ¥ç¶šã‚’ã—ã«ããã™ã‚‹ã®ãŒã€SYN Floodæ”»æ’ƒã€‚ SYN cookiesã¨ã¯ã€ã“ã®ã€ŒSYN Floodã€æ”»æ’ƒã¸ã®å¯¾ç–ã®ãŸã‚ã®ä»•çµ„ã¿ã§ã™ã€‚ ã¨ã„ã†ã‚ã‘ã§ä»¥ä¸‹ã€Linuxã§ã®ã“ã®å®Ÿè£…ã«é–¢ã—ã¦ã‚³ã‚¢ãªæ„Ÿã˜ã€‚ syncookiesã¯ã€åˆæœŸã‚·ãƒ¼ã‚±ãƒ³ã‚¹ç•ªå·ã‚„ã€MSSç‰ã‚’ã‚¨ãƒ³ã‚³ãƒ¼ãƒ‰ã—ã€ä¸€æ–¹å‘ãƒãƒƒã‚·ãƒ¥é–¢æ•°ã§ã‚»ã‚ãƒ¥ãƒªãƒ†ã‚£ã‚’é«˜ã‚ã¦ã€æŽ¥ç¶šè¦æ±‚å…ƒã«è¿”ã™SYN/ACKãƒ‘ã‚±ãƒƒãƒˆã®ã‚·ãƒ¼ã‚±ãƒ³ã‚¹ç•ªå·éƒ¨åˆ†ã«ã¤ã£ã“ã‚“ã§ã€æŽ¥ç¶šå¾…ã¡ã‚ãƒ¥ãƒ¼ã‚’æ¶ˆè²»ã—ãªãã¦æ¸ˆã‚€ã‚ˆã†ã«ã™ã‚‹ä»•çµ„ã¿ã€‚ syn queueãŒã„ã£ã±ã„ã«ãªã£ãŸå ´åˆï¼Œ syncookiesãŒç„¡åŠ¹ã§ã‚ã‚Œã°dropã—ã¦ã—ã¾ã†ã¨ã“ã‚ã‚’ï¼Œ syn cookieã¤ãã®SYNACKã‚’
y_uuki 2015/03/16
linux

TCP
ãƒªãƒ³ã‚¯
Programming UNIX Sockets in C - Frequently Asked Questions: ã‚¯ãƒ©ã‚¤ã‚¢ãƒ³ãƒˆã¨ã‚µãƒ¼ãƒ(TCP/SOCK_STREAM)ä¸¡æ–¹ã«é–¢ã™ã‚‹è³ªå•
Previous Next Table of Contents 2. ã‚¯ãƒ©ã‚¤ã‚¢ãƒ³ãƒˆã¨ã‚µãƒ¼ãƒ(TCP/SOCK_STREAM)ä¸¡æ–¹ã«é–¢ã™ã‚‹è³ªå• 2.1 ç›¸æ‰‹å´ã®ã‚½ã‚±ãƒƒãƒˆãŒé–‰ã˜ã‚‰ã‚ŒãŸã“ã¨ã‚’ã©ã†ã‚„ã£ã¦çŸ¥ã‚‹ã“ã¨ãŒã§ãã¾ã™ã‹? Andrew Gierth æ° ( [email protected]emon.co.uk) ã‚ˆã‚Š: ç§ã®çŸ¥ã‚‹é™ã‚Šâ€¦ ç›¸æ‰‹å´ãŒ (SO_LINGER ã‚’ä½¿ã£ãŸã‚„ã‚„ã“ã—ã„ã“ã¨ã‚’ã—ãªã„ã§) close() ã™ã‚‹ã‹çµ‚äº†ã—ãŸã¨ã™ã‚‹ã¨ã€ã“ã¡ã‚‰ã® read() ã®å‘¼ã³ å‡ºã—ã¯ 0 ã‚’è¿”ã™ã¯ãšã§ã™ã€‚åŒã˜å ´åˆã§ã€write() å‘¼ã³å‡ºã—ã§ä½•ãŒ èµ·ã“ã‚‹ã‹ã¯ã€ã‚‚ã†ã¡ã‚‡ã£ã¨ã‚ã‹ã‚Šã¥ã‚‰ã„ã§ã™ã€‚ç›´å¾Œã®å‘¼ã³å‡ºã—æ™‚ã§ã¯ãª ãã€ãã®æ¬¡ã®å‘¼ã³å‡ºã—æ™‚ã«EPIPE ãŒè¿”ã‚‹ã§ã—ã‚‡ã†ã€‚ ã‚‚ã—ç›¸æ‰‹ãŒå†èµ·å‹•ã™ã‚‹ã‹ l_onoff = 1, l_linger = 0 ã‚’è¨å®šã—ã¦ ã‹ã‚‰é–‰ã˜ãŸã¨ã™ã‚‹ã¨ã€read() ã‹ã‚‰ã¯(
y_uuki 2015/02/16
TCP

linux
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‚¯
Linux ftrace TCP Retransmit Tracing
Recent posts: 29 Oct 2024 Â» AI Flame Graphs 22 Jul 2024 Â» No More Blue Fridays 24 Mar 2024 Â» Linux Crisis Tools 17 Mar 2024 Â» The Return of the Frame Pointers 10 Mar 2024 Â» eBPF Documentary 28 Apr 2023 Â» eBPF Observability Tools Are Not Security Tools 01 Mar 2023 Â» USENIX SREcon APAC 2022: Computing Performance: What's on the Horizon 17 Feb 2023 Â» USENIX SREcon APAC 2023: CFP 02 May 2022 Â» Brendan
y_uuki 2014/09/14
linux

tcp
ãƒªãƒ³ã‚¯
listen()ã®backlogãŒä¸è¶³ã—ãŸéš›ã®TCP_DEFER_ACCEPTã®å‹•ä½œã«ã¤ã„ã¦ - blog.nomadscafe.jp
TCP_DEFER_ACCEPTã¯ã€Linuxã§ã‚µãƒãƒ¼ãƒˆã•ã‚Œã¦ã„ã‚‹TCPã®ã‚ªãƒ—ã‚·ãƒ§ãƒ³ã§ã€ã‚µãƒ¼ãƒå´ã§ä½¿ç”¨ã—ãŸå ´åˆã«ã¯accept(2)ã‹ã‚‰ã®ãƒ–ãƒãƒƒã‚¯è§£é™¤ã‚’TCPæŽ¥ç¶šãŒå®Œäº†ã—ãŸã‚¿ã‚¤ãƒŸãƒ³ã‚°ã§ã¯ãªãæœ€åˆã®ãƒ‡ãƒ¼ã‚¿ãŒåˆ°ç€ã—ãŸã‚¿ã‚¤ãƒŸãƒ³ã‚°ã§è¡Œã£ã¦ãã‚Œã‚‹ã‚ªãƒ—ã‚·ãƒ§ãƒ³ã§ã™ã€‚ Webã‚µãƒ¼ãƒãƒ»ã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã‚µãƒ¼ãƒã§ã¯ãƒªã‚¯ã‚¨ã‚¹ãƒˆãŒåˆ°ç€ã—ã¦ã‹ã‚‰accept(2)ã®ãƒ–ãƒãƒƒã‚¯ã‚’è§£é™¤ã™ã‚‹ã®ã§ã€ãƒªã‚¯ã‚¨ã‚¹ãƒˆã®åˆ°ç€ã‚’Webã‚µãƒ¼ãƒãƒ»ã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã‚µãƒ¼ãƒã§å¾…ã¤å¿…è¦ãŒãªããªã‚Šã€ç‰¹ã«preforkåž‹ã®ã‚µãƒ¼ãƒã§ã¯åŠ¹çŽ‡çš„ã«ãƒ—ãƒã‚»ã‚¹ã‚’ä½¿ãˆã‚‹ã‚ˆã†ã«ãªã‚‹ã¨ã„ã†åˆ©ç‚¹ãŒã‚ã‚Šã¾ã™ã€‚Perlã§ã¯StarletãŒã“ã®æ©Ÿèƒ½ã‚’æœ‰åŠ¹ã«ã—ã¦ã„ã¾ã™ ã¨ã“ã‚ãŒã€æŸã‚µãƒ¼ãƒ“ã‚¹ã§TCP_DEFER_ACCEPTãŒæœ‰åŠ¹ã«ã‚‚é–¢ã‚ã‚‰ãšã€acceptå¾Œã®readã§ãƒ‡ãƒ¼ã‚¿ãŒèªã‚ãšã€æœ€æ‚ªã®å ´åˆã€ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆã®timeoutã§ã‚ã‚‹5åˆ†é–“ãƒ—ãƒã‚»ã‚¹ãŒã‚¹ãƒˆãƒ¼ãƒ«ã™ã‚‹ã“ã¨ãŒã‚ã‚Šã¾ã—ãŸã€‚strace
y_uuki 2014/07/17
TCP

linux

kernel
ãƒªãƒ³ã‚¯
Linuxã§TCP_DEFER_ACCEPTãŒæœ‰åŠ¹ã§ã‚‚acceptå¾Œreadã§ããªã„ç†ç”±
listen()ã®backlogãŒä¸è¶³ã—ãŸéš›ã®TCP_DEFER_ACCEPTã®å‹•ä½œã«ã¤ã„ã¦ - blog.nomadscafe.jpã¨ã„ã†è¨˜äº‹ã®ä¸ã§ã€listen backlog ãŒã‚ãµã‚ŒãŸå¾Œã« accept(2) ã™ã‚‹ã¨ã€ãã®å¾Œã® read(2) ãŒ EAGAIN ã‚’è¿”ã—ãŸã‚Šã€æŽ¥ç¶šãŒä¸å®‰å®šã«ãªã‚‹ã¨ã„ã†äº‹è±¡ãŒèª¬æ˜Žã•ã‚Œã¦ã„ã¾ã—ãŸã€‚æ°—ã«ãªã£ãŸã®ã§èª¿ã¹ã¦ã¿ãŸã“ã¨ã‚’ã¾ã¨ã‚ã¾ã™ã€‚ çµè«–ã‹ã‚‰è¨€ã†ã¨ã“ã‚Œã¯Linuxã®ä»•æ§˜ã§ã™ã€‚manã®tcp(7)ã‚’è¦‹ã‚‹ã¨ã€ TCP_DEFER_ACCEPT (since Linux 2.4) Allow a listener to be awakened only when data arrives on the socket. Takes an integer value (seconds), this can bound the maximum number of
y_uuki 2014/05/09
TCP

linux

kernel
ãƒªãƒ³ã‚¯
ã€Œnet.ipv4.tcp_tw_recycleã€ã‚’æœ‰åŠ¹ã«ã™ã‚‹ã®ã¯ï¼ˆå ´åˆã«ã‚ˆã£ã¦ã¯ï¼‰ã‚„ã‚ãŸæ–¹ãŒã„ã„ - pullphone's blog
ãã‚‚ãã‚‚ã€Œtcp_tw_recycleã€ã£ã¦ãªã«ï¼Ÿ TIME_WAITçŠ¶æ…‹ã®ã‚½ã‚±ãƒƒãƒˆ*1ã‚’é«˜é€Ÿã«å†åˆ©ç”¨ã™ã‚‹ãŸã‚ã®Linuxã‚«ãƒ¼ãƒãƒ«ç‰¹æœ‰ã®ä»•çµ„ã¿ã‚‰ã—ã„ã€‚ ã€Œ/etc/sysctl.confã€ã§ã“ã„ã¤ï¼ˆnet.ipv4.tcp_tw_recycleï¼‰ã‚’1ã«ã—ã¦ã‚„ã£ã¦ã€Œsysctl -pã€ã™ã‚‹ã ã‘ã§æœ‰åŠ¹ã«ãªã‚Šã¾ã™ã€‚ ã€ŒTIME_WAITã®ã‚½ã‚±ãƒƒãƒˆã‚’å°‘ãªãã—ã¦ãã‚Œã‚‹ã‚“ã§ã—ã‚‡ï¼Ÿã„ã„ã˜ã‚ƒã‚“ï¼ã€ã¨ã„ã†æ„Ÿã˜ã§è¨å®šã—ã¦ã—ã¾ã„ãã†ã«ãªã‚Šã¾ã™ãŒãƒ»ãƒ»ãƒ» ãªã‚“ã§ãƒ€ãƒ¡ãªã®ï¼Ÿ çµè«–ã‹ã‚‰è¨€ã†ã¨ã€ã“ã„ã¤ã‚’æœ‰åŠ¹ã«ã—ãŸã¨ãã€åŒã˜ã‚°ãƒãƒ¼ãƒãƒ«IPã®ã‚¯ãƒ©ã‚¤ã‚¢ãƒ³ãƒˆã‹ã‚‰ã®æŽ¥ç¶šã§ã‹ã¤TCPãƒ‘ã‚±ãƒƒãƒˆã«ã‚¿ã‚¤ãƒ ã‚¹ã‚¿ãƒ³ãƒ—æƒ…å ±ãŒå…¥ã£ã¦ã„ã‚‹å ´åˆã§ã€ã»ã¼åŒæ™‚ã«ãƒ‘ã‚±ãƒƒãƒˆã‚’é€ã‚‹ã¨ã€å¤ã„ã‚¿ã‚¤ãƒ ã‚¹ã‚¿ãƒ³ãƒ—ã®æ–¹ã®ãƒ‘ã‚±ãƒƒãƒˆã‚’å‹æ‰‹ã«ãƒ‰ãƒãƒƒãƒ—ã—ã¡ã‚ƒã„ã¾ã™ã€‚ ã©ã†ã„ã†ã“ã¨ãªã®ãƒ»ãƒ»ãƒ» ã€Œtcp_tw_recycleã€ã¯ã€ã€ŒåŒä¸€IPã‹ã‚‰ã®ãƒ‘ã‚±ãƒƒãƒˆãŒåˆ°ç€ã—ãŸã¨ãã€ä½¿ã£ã¦ã„ãŸã‚½ã‚±
y_uuki 2014/02/04
linux

TCP

kernel
ãƒªãƒ³ã‚¯
1 2 3 æ¬¡ã®ãƒšãƒ¼ã‚¸