『「ペパボ・ã¯ã¦ãªæŠ€è¡“大会 〜インフラ技術基盤〜ã€ã‚’ç¦å²¡ã¨äº¬éƒ½ã§é–‹å‚¬ï¼ - Hatena Developer Blogã€ã¸ã®ã‚³ãƒ¡ãƒ³ãƒˆ
ã“ã‚“ã«ã¡ã¯ã€ã¯ã¦ãªã‚¦ã‚§ãƒ–オペレーションエンジニア㮠id:y_uuki ã§ã™ã€‚ GMOãƒšãƒ‘ãƒœæ ªå¼ä¼šç¤¾ã•ã‚“ã¨ã®å…±åŒã§ã€Œãƒšãƒ‘ボ・ã¯ã¦ãªæŠ€è¡“大会ã€ã‚’ç¦å²¡ã¨äº¬éƒ½ã§é–‹å‚¬ã—ã¾ã™ã€‚ 今回ã¯ã€ãƒšãƒ‘ボã•ã‚“ã¨ã¯ã¦ãªã®ã‚¤ãƒ³ãƒ•ãƒ©ã‚¨ãƒ³ã‚¸ãƒ‹... 112 人ãŒãƒ–ックマーク・24 件ã®ã‚³ãƒ¡ãƒ³ãƒˆ
「ペパボ・ã¯ã¦ãªæŠ€è¡“大会 〜インフラ技術基盤〜ã€ã‚’ç¦å²¡ã¨äº¬éƒ½ã§é–‹å‚¬ï¼ - Hatena Developer Blog
ã“ã‚“ã«ã¡ã¯ã€ã¯ã¦ãªã‚¦ã‚§ãƒ–オペレーションエンジニア㮠id:y_uuki ã§ã™ã€‚ GMOãƒšãƒ‘ãƒœæ ªå¼ä¼šç¤¾ã•ã‚“ã¨ã®å…±åŒã§ã€Œãƒšãƒ‘ボ・ã¯ã¦ãªæŠ€è¡“大会ã€ã‚’ç¦å²¡ã¨äº¬éƒ½ã§é–‹å‚¬ã—ã¾ã™ã€‚ 今回ã¯ã€ãƒšãƒ‘ボã•ã‚“ã¨ã¯ã¦ãªã®ã‚¤ãƒ³ãƒ•ãƒ©ã‚¨ãƒ³ã‚¸ãƒ‹ã‚¢ãƒ»ã‚¦ã‚§ãƒ–オペレーションエンジニアãŒé›†ã¾ã‚Šã€ã‚¤ãƒ³ãƒ•ãƒ©æŠ€è¡“基盤をテーマã¨ã—ãŸæ§˜ã€…ãªãƒˆãƒ”ックã«ã¤ã„ã¦ã”紹介ã™ã‚‹äºˆå®šã§ã™ã€‚ 「Webサービスã®é‹ç”¨ç®¡ç†ã‹ã‚‰ã®è§£æ”¾ã€ã€ã€ŒSTNSã§å®Ÿç¾ã™ã‚‹æ¬¡ä¸–代ユーザー管ç†ã€ã€ã€Œã‚·ã‚¹ãƒ†ãƒ を改善ã™ã‚‹è¡Œå‹•ãƒã‚° ã€ã€ã€ŒWebサービスã«ãŠã‘ã‚‹OSã¨ãƒŸãƒ‰ãƒ«ã‚¦ã‚§ã‚¢ã®ãƒ‘フォーマンス改善ã€ã¨ã„ã£ãŸãƒ†ãƒ¼ãƒžã«ã‚ˆã‚‹ãƒˆãƒ¼ã‚¯ã‚’予定ã—ã¦ã„ã¾ã™ã€‚ ã•ã‚‰ã«ã€ãƒšãƒ‘ボã•ã‚“ã¨ã¯ã¦ãªã®è‹¥æ‰‹ã‚¤ãƒ³ãƒ•ãƒ©ã‚¨ãƒ³ã‚¸ãƒ‹ã‚¢ãƒ»ã‚¦ã‚§ãƒ–オペレーションエンジニアã«ã‚ˆã‚‹åº§è«‡ä¼šã‚‚予定ã—ã¦ã„ã¾ã™ã€‚ 日程ã¨å ´æ‰€ ã¯ã¦ãªãƒ»ãƒšãƒ‘ボ技術大会: 7月2æ—¥(土)@ã¯ã¦ãªäº¬éƒ½ã‚ªãƒ•ã‚£ã‚¹ ペパボ・ã¯ã¦ãªæŠ€è¡“大会: 7月9æ—¥(土)@ペパボç¦å²¡æ”¯ç¤¾
nginx ã¡ã‚‡ã£ã¨ä¸æ€è°ã ã£ãŸãƒªã‚¯ã‚¨ã‚¹ãƒˆãƒªãƒˆãƒ©ã‚¤ã®ãŠè©± - Cybozu Inside Out | サイボウズエンジニアã®ãƒ–ãƒã‚°
ã“ã‚“ã«ã¡ã¯ã€Sustain ãƒãƒ¼ãƒ ã®å±±å£ã§ã™ã€‚ 今サイボウズリモートサービスã¨ã„ã†VPN・ä¸ç¶™ã‚µãƒ¼ãƒ“スã§ä½¿ç”¨ã™ã‚‹ L7LB ã‚’ nginx ã«ç§»è¡Œã—よã†ã¨ã„ã‚ã„ã‚調査をã—ã¦ã„ã¾ã™ã€‚ nginx をリãƒãƒ¼ã‚¹ãƒ—ãƒã‚ã‚·ã¨ã—ã¦ä½¿ç”¨ã™ã‚‹éš›ã«å°‘々障害ã¨ãªã‚‹å‹•ä½œãŒã‚ã‚Šã€ãã‚Œã«é–¢ã™ã‚‹ã‚¢ãƒƒãƒ—デートãŒæœ€è¿‘ã‚ã£ãŸã®ã§ã€ä»Šå›žã¯ãã®ãŠè©±ã§ã™ã€‚ 3è¡Œã§åˆ†ã‹ã‚‹æœ¬è¨˜äº‹ã®å†…容 nginx をリãƒãƒ¼ã‚¹ãƒ—ãƒã‚ã‚·ã¨ã—ã¦ä½¿ã†å ´åˆã€ãƒªã‚¯ã‚¨ã‚¹ãƒˆãŒãƒãƒƒã‚¯ã‚¨ãƒ³ãƒ‰ã‚µãƒ¼ãƒã«äºŒé‡ã§é£›ã¶å¯èƒ½æ€§ãŒã‚ã‚‹ 対ç–ã¨ã—㦠proxy_request_buffering off ãŒä½¿ãˆãã†ã ãŒã€ä½¿ãˆãªã„ nginx 1.9.13 ã§ä»•æ§˜å¤‰æ›´ãŒå…¥ã‚Šã€POST, LOCK, PATCH メソッドã®ãƒªã‚¯ã‚¨ã‚¹ãƒˆã¯äºŒé‡ã§é£›ã°ãªã„ã“ã¨ãŒä¿è¨¼ã•ã‚ŒãŸ nginx ã®å—動的㪠health check ã¨ãƒªã‚¯ã‚¨ã‚¹ãƒˆã®ãƒªãƒˆãƒ©ã‚¤å‹•ä½œ nginx ã¯ãƒªãƒãƒ¼ã‚¹ãƒ—ãƒã‚ã‚·ã¨ã—ã¦ä½¿ã†å ´åˆã«
『プãƒã‚°ãƒ©ãƒŸãƒ³ã‚°è¨€èªžGoã€ç¿»è¨³ä½œæ¥ãŒçµ‚了ã—ã¾ã—ãŸ
プãƒã‚°ãƒ©ãƒŸãƒ³ã‚°è¨€èªžGo 作者: Alan A.A. Donovan出版社/メーカー: 丸善出版発売日: 2016/06/15メディア: å˜è¡Œæœ¬ï¼ˆã‚½ãƒ•ãƒˆã‚«ãƒãƒ¼ï¼‰ 2015å¹´4月20æ—¥ã‹ã‚‰åŽŸè‘—ã®åŽŸç¨¿ã®ãƒ¬ãƒ“ューを始ã‚ã¦ã€ãƒ¬ãƒ“ューãŒçµ‚ã‚ã£ãŸã®ãŒ2015å¹´9月5æ—¥ã§ã—ãŸã€‚ãã®é–“ã«ã€äºŒå›žèªã¿è¿”ã—ã¦ã„ã¾ã™ã€‚翻訳ã¯2015å¹´9月13æ—¥ã«ç€æ‰‹ã—始ã‚ã¦ã€ç§ã®ã™ã¹ã¦ã®ä½œæ¥ãŒæ˜¨æ—¥ï¼ˆ2016å¹´5月8日）終了ã—ã¾ã—ãŸã€‚今日ã€å…¥ç¨¿ã¨ãªã‚Šå°åˆ·æ‰€ã§ã®å°åˆ·ã®æº–å‚™ãŒå§‹ã¾ã‚Šã¾ã™ã®ã§ã€äºˆå®šé€šã‚Š6月15日刊行ã¨ãªã‚Šã¾ã™ã€‚ 今回ã¯ã€åŽŸè‘—ã®ãƒ¬ãƒ“ューã«98時間ã€ç¿»è¨³ã«473時間を費やã—ãŸã“ã¨ã«ãªã‚Šã¾ã™ã€‚ã¡ãªã¿ã«ã€ŽAPIデザインã®æ¥µæ„ã€ã¯ã€ç¿»è¨³ã ã‘ã§ã—ãŸãŒ500時間を費やã—ã¦ã„ã¾ã™ã€‚ã“ã‚Œã¯ã€ç§ãŒç§çš„時間ã«è²»ã‚„ã—ãŸæ™‚é–“ã§ã‚ã‚Šã€å‡ºç‰ˆç¤¾ã®æ‹…当者やレビューアã«ã‚ˆã‚‹ãƒ¬ãƒ“ãƒ¥ãƒ¼ã‚’åŠ ãˆã‚‹ã¨å¤šãã®æ™‚é–“ãŒè²»ã‚„ã•ã‚ŒãŸã“ã¨ã«ãªã‚Šã¾ã™ã€‚ã¾ãŸã€ç¤¾å†…ã®Goè¨€èªžç ”ä¿®ã§ã¯ç¿»
Golangã®GCを追ã†
Go1.5ã¨Go1.6ã§Goã®GCã®ãƒ¬ã‚¤ãƒ†ãƒ³ã‚·ãŒå¤§ãã改善ã•ã‚ŒãŸï¼Žã“ã®å¤‰æ›´ã«ã¤ã„ã¦ã€Œã¡ã‚ƒã‚“ã¨ã€ç†è§£ã™ã‚‹ãŸã‚，アルゴリズムレベルã§Goã®GCã«ã¤ã„ã¦è¿½ã£ã¦ã¿ãŸï¼Ž ã¾ãšGoã®GCã®ç¾çŠ¶ã‚’パフォーマンス（レイテンシ）ã®è¦³ç‚¹ã‹ã‚‰ã¾ã¨ã‚る．次ã«å…·ä½“çš„ãªã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã«ã¤ã„ã¦ï¼Œãã—ã¦æœ€å¾Œã«å®Ÿéš›ã®ç¾å ´ã§ã®ãƒãƒ¥ãƒ¼ãƒ‹ãƒ³ã‚°ã¯ã©ã†ã™ã‚Œã°è‰¯ã„ã®ã‹ã«ã¤ã„ã¦è§£èª¬ã™ã‚‹ï¼Ž Goã®GCã®ä»Š 最åˆã«Goã®GCã®æœ€è¿‘ã®æµã‚Œï¼ˆ2016å¹´5月ã¾ã§ï¼‰ã‚’ã¾ã¨ã‚る． Go1.4ã¾ã§ã¯å˜ç´”ãªStop The World（STW）GCãŒå®Ÿè£…ã•ã‚Œã¦ã„ãŸãŒGo1.5ã‹ã‚‰ã¯æ–°ãŸãªGCアルゴリズムãŒå°Žå…¥ã•ã‚ŒãŸï¼Žå°Žå…¥ã®éš›ã«è¨å®šã•ã‚ŒãŸæ•°å€¤ç›®æ¨™ã¯å¤§ããªãƒ’ープサイズã«ãŠã„ã¦ã‚‚レイテンシを10ms以下ã«æŠ‘ãˆã‚‹ã“ã¨ã§ã‚ã£ãŸï¼ŽGo1.5ã§æ–°ãŸãªã‚¢ãƒ«ã‚´ãƒªãƒ ãŒå®Ÿè£…ã•ã‚ŒGo1.6ã§æœ€é©åŒ–ãŒè¡Œã‚ã‚ŒãŸï¼Ž 以下ã¯å…¬é–‹ã•ã‚Œã¦ã„るベンãƒãƒžãƒ¼ã‚¯ï¼Žã¾ãšã¯Go1.5を見る． Gophe
Design Tradeoffs for Data Deduplication Performance in Backup Workloads | USENIX
Min Fu, Dan Feng, and Yu Hua, Huazhong University of Science and Technology; Xubin He, Virginia Commonwealth University; Zuoning Chen, National Engineering Research Center for Parallel Computer; Wen Xia and Yucheng Zhang, Huazhong University of Science and Technology; Yujuan Tan, Chongqing University Data deduplication has become a standard component in modern backup systems. In order to understan
Non-blocking Writes to Files | USENIX
Daniel Campello and Hector Lopez, Florida International University; Luis Useche, Google Inc.; Ricardo Koller, IBM T. J. Watson Research Center; Raju Rangaswami, Florida International University Writing data to a page not present in the file-system page cache causes the operating system to synchronously fetch the page into memory first. Synchronous page fetch defines both policy (when) and mechanis
Failure-Atomic Updates of Application Data in a Linux File System | USENIX
Rajat Verma and Anton Ajay Mendez, Hewlett-Packard; Stan Park, Hewlett-Packard Labs; Sandya Mannarswamy, Hewlett-Packard; Terence Kelly and Charles B. Morrey III, Hewlett-Packard Labs We present the design, implementation, and evaluation of a file system mechanism that protects the integrity of application data from failures such as process crashes, kernel panics, and power outages. A simple inter
Flash Reliability in Production: The Expected and the Unexpected | USENIX
As solid state drives based on flash technology are becoming a staple for persistent data storage in data centers, it is important to understand their reliability characteristics. While there is a large body of work based on experiments with individual flash chips in a controlled lab environment under synthetic workloads, there is a dearth of information on their behavior in the field. This paper
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 Technology (KAIST); Sunghwan Ihm, Princeton University; Dongsu Han and KyoungSoo Park, Korea Advanced Institute of Science and Technology (KAIST) Community Award Scaling the performance of short TCP connections on multicore systems is fundamentally challenging. Although many proposals have attem
SF-TAP: Scalable and Flexible Traffic Analysis Platform Running on Commodity Hardware | USENIX
"LISA is the best mix of training, talks, and networking of any events I've been to. That's the reason I've been to 11 of them.†LISA14 Attendee "In a world where technology changes rapidly, it's hard to find great resources for in-depth expertise. LISA has a terrific combination of people who create the technology and people who implement it effectively for their companies." Matt Simmons, Northea
リリースã€éšœå®³æƒ…å ±ãªã©ã®ã‚µãƒ¼ãƒ“スã®ãŠçŸ¥ã‚‰ã›
最新ã®äººæ°—エントリーã®é…ä¿¡
j次ã®ãƒ–ックマーク
kå‰ã®ãƒ–ックマーク
lã‚ã¨ã§èªã‚€
eコメント一覧を開ã
oページを開ã
F2FS: A New File System for Flash Storage | USENIX
Access Characteristic Guided Read and Write Cost Regulation for Performance Improvement on Flash Memory | USENIX
BTrDB: Optimizing Storage System Design for Timeseries Processing | USENIX
Thread and Memory Placement on NUMA Systems: Asymmetry Matters | USENIX
Optimizing Every Operation in a Write-optimized File System | USENIX
ParaSwift: File I/O Trace Modeling for the Future | USENIX
The Truth About MapReduce Performance on SSDs | USENIX
Automatic and Dynamic Configuration of Data Compression for Web Servers | USENIX
Realtime High-Speed Network Traffic Monitoring Using ntopng | USENIX
Spyglass: Demand-Provisioned Linux Containers for Private Network Access | USENIX
Dynamic Provisioning of Storage Workloads | USENIX
{{#tags}}- {{label}}
{{/tags}}