Lessons Learned in Concurrency with Ruby - Part I
The Engine Yard PaaS products are highly complex, and span thousands of servers to serve our customers. Ensuring that our applications run quickly and reliably requires a good understanding of concurrency and parallelization. This blog post is a recap of my journey down the rabbit hole of concurrency in an attempt to solve this problem in my own development efforts. Part II will be the continuatio
Go言語ã§CPUæ•°ã«å¿œã˜ã¦ä¸¦åˆ—処ç†æ•°ã‚’制é™ã™ã‚‹
è² è·ã®ã‹ã‹ã‚‹å‡¦ç†ã‚’制é™ãªã—ã«ä¸¦åˆ—化ã—ã¦ã‚‚æ„味ãªã„．処ç†ã®ä¸¦åˆ—数を予測å¯èƒ½ãªå ´åˆã¯ï¼Œå½“ãŸã‚Šã‚’ã¤ã‘ã¦æœ€é©åŒ–ã™ã‚‹ã®ã‚‚よã„ãŒï¼Œä¸æ˜Žç¢ºãªå ´åˆã¯ï¼ŒCPUæ•°ã«ã‚ˆã‚‹åˆ¶é™ãŒå˜ç´”ãª1ã¤ã®è§£ã«ãªã‚‹ï¼Ž TL;DR CPUæ•°ã«å¿œã˜ãŸãƒãƒƒãƒ•ã‚¡é•·ã®Channelを使ã£ã¦ã‚»ãƒžãƒ•ã‚©ã‚’実装ã™ã‚‹ï¼Ž 実例 mitchellh/gox goxã¯Go言語製ã®ãƒ„ールを並列コンパイルã™ã‚‹ãƒ„ール．コンパイルã®å‡¦ç†ã¯é‡ã„ãŸã‚，デフォルトã§ä¸¦åˆ—処ç†æ•°ã‚’CPUæ•°ã§åˆ¶é™ã—ã¦ã„る． ç°¡å˜ãªä¾‹ 例ãˆã°ï¼Œä»¥ä¸‹ã®ã‚ˆã†ãªå˜ç´”ãªä¸¦åˆ—処ç†ã‚’考ãˆã‚‹ï¼Žheavy()（é‡ã„処ç†ï¼‰ã‚’並列ã§å®Ÿè¡Œã™ã‚‹ï¼Ž package main import ( "fmt" "sync" "time" ) func heavy(i int) { fmt.Println(i) time.Sleep(5 * time.Second) } func main() { var wg sync.W
Clojure プãƒã‚°ãƒ©ãƒ を並列化ã—ãŸã‚‰é…ããªã£ãŸã‚Šè¤‡é›‘ã«ãªã£ãŸã‚Šã™ã‚‹ã¨ã„ã†ã®ã¯ä½•ã‹ã®é–“é•ã„ã§ã¯ãªã„ã‹ã¨æ€ã£ã¦ç¢ºã‹ã‚ãŸã‚‰ï¼Œã‚„ã£ã±ã‚Šé–“é•ã„ã ã£ãŸä»¶ - tnoda-clojure
å‰å›žã® Java ã¨æ¯”ã¹ã¦ Clojure ㌠3 å€ä»¥ä¸Šé…ã„ã¨ã„ã†ã®ã¯ä½•ã‹ã®é–“é•ã„ã§ã¯ãªã„ã‹ã¨æ€ã£ã¦ç¢ºã‹ã‚ãŸã‚‰ï¼Œã‚„ã£ã±ã‚Šé–“é•ã„ã ã£ãŸä»¶ ã®ç¶šãã§ã™ï¼Ž 今回ã¯ï¼Œå‰å›žæ™®é€šã«æ›¸ã„ãŸä¸¦åˆ—化ãªã— Clojure プãƒã‚°ãƒ©ãƒ を並列化ã•ã›ã¾ã™ï¼Ž 普通ã«æ›¸ã„ãŸä¸¦åˆ—化ãªã— Clojure プãƒã‚°ãƒ©ãƒ ã®ã‚³ãƒ¼ãƒ‰ã¯å‰ã®è¨˜äº‹ã‚’å‚ç…§ã—ã¦ãã ã•ã„． 「普通ã«æ›¸ã„㟠Clojure プãƒã‚°ãƒ©ãƒ ã€ã®ä¸¦åˆ—化ã“ã®è¨˜äº‹ã§ã¯ï¼Œ pmapReducersã® 2 ã¤ã®æ–¹æ³•ã‚’使ã£ã¦ã€Œæ™®é€šã«æ›¸ã„㟠Clojure プãƒã‚°ãƒ©ãƒ ã€ã‚’並列化ã—， 並列化å‰ã¨ä¸¦åˆ—化後ã®å®Ÿè¡Œæ™‚間を比較ã—，ã©ã¡ã‚‰ãŒé€Ÿã„ã‹ã‚’確èªã—ã¾ã™ï¼Ž ã¾ãŸï¼Œã‚³ãƒ¼ãƒ‰ã®å¤‰æ›´ç‚¹ã‚’確èªã™ã‚‹ã“ã¨ã§ï¼Œä¸¦åˆ—化ã«ã‚ˆã‚Šã©ã‚Œã ã‘コードãŒè¤‡é›‘化ã™ã‚‹ã‹ã‚’è°è«–ã—ã¾ã™ï¼Ž pmap ã«ã‚ˆã‚‹ä¸¦åˆ—化「普通ã«æ›¸ã„㟠Clojure プãƒã‚°ãƒ©ãƒ ã€ã‹ã‚‰ã®å¤‰æ›´ç‚¹ã¯æ¬¡ã® 2 ã¤ã§ã™: map ã‚’ pmap ã«ç½®ãæ›ãˆãŸï¼ˆ1
JVisualVM and Clojure
JVisualVM is a Java Virtual machine monitoring tool that allows you to see a live view of the Java memory, cpu and threads that are currently active. In this post, I'll look at how easy it is to profile Clojure code using JVisualVM. First step is to get jvisualvm installed. If you've got a recent JDK then it's already available in the bin directory of the JDK install. If not, then upgrade to the l
Parallel computing - Wikipedia
1.3 Race conditions, mutual exclusion, synchronization, and parallel slowdown
concurrentã¨parallelã®é•ã„説明会
ã„ãã¤ã‹ä¾‹ã‚’挙ã’ãªãŒã‚‰ã‚ˆã話題ã«ãªã‚‹concurrent(並行)ã¨parallel(並列)ã®é•ã„ã‚’n_sodaã•ã‚“ã¨è©±ã—åˆã„ã¾ã—ãŸã€‚ concurrentã¨parallelã®é•ã„ãŒæŽ˜ã‚Šä¸‹ã’られã¦æ”¹ã‚ã¦å‹‰å¼·ã«ãªã‚Šã¾ã—ãŸã€‚
並行処ç†ã«ã¤ã„ã¦ã®å€‹äººçš„æ€è€ƒã®ã¾ã¨ã‚　-処ç†ã®åŒæœŸåŒ–ç·¨- - Akinekoã®æ—¥è¨˜
最近転è·æ´»å‹•ã®ä¸€ã¤ã¨ã—ã¦ä½•ã‹å½¢ã«ãªã‚‹ã‚‚ã®ã‚’作ã‚ã†ã¨ãƒ„イッター関係ã®ã‚¢ãƒ—リを作æˆã—ã¦ã„ã‚‹ã®ã§ã™ãŒã€ãã®éš›ã«ä¸¦è¡Œå‡¦ç†åŒ–ã—ãŸã„部分ãŒå‡ºã¦ãã¦ã‚ã‚Œã“ã‚Œã¨è€ƒãˆãŸã®ã§ã€ãã®è€ƒãˆã‚’æ•´ç†ã™ã‚‹ç‚ºã«ã‚‚ã¾ã¨ã‚を書ããŸã„ã¨æ€ã„ã¾ã™ã€‚コード例ã¯å…¨ã¦Javaを使用ã—ã¦ã„ã¾ã™ãŒæœ€å¾Œã«ä»–言語ã®æ©Ÿèƒ½ã‚‚紹介ã—ã¦ã„ã¾ã™ã—ã€è€ƒãˆæ–¹è‡ªä½“ã¯ä»–ã®è¨€èªžã«ã‚‚é©ç”¨ã§ãã‚‹ã‹ã¨æ€ã„ã¾ã™ã€‚ ã¾ãŸã€æ–‡ç« ã®æ›¸ãæ–¹ãŒå®šã¾ã£ã¦ã„ãªã„ã®ã§é€”ä¸ã§æ–‡ç« ãŒå¤‰ã‚ã£ãŸã‚Šã¨èªã¿ã«ãã‹ã£ãŸã‚‰ã™ã¿ã¾ã›ã‚“。一応ã€é€±ä¸€ã§æ›´æ–°ã™ã‚‹ã“ã¨ã‚’先週辺りã‹ã‚‰å§‹ã‚ã¾ã—ãŸã®ã§ã€ãã®å†…安定ã™ã‚‹ã¨æ€ã„ã¾ã™ã‹ã‚‰ãã‚Œã¾ã§æˆ‘æ…¢ã—ã¦ãŠå¾…ã¡ãã ã•ã„ï½— 書ã‹ãªãゃã„ã‘ãªã„記事ã®å®¿é¡Œã‚„自分ã®çŸ¥è˜ã‚‚ã‚る程度溜ã¾ã£ã¦ãã¾ã—ãŸã—ã＾＾； ã€è¿½è¨˜ã€‘ ã¾ã¨ã‚→並行処ç†ã®åŠ©ã‘ã¨ãªã‚‹è€ƒãˆã¸å¤‰æ›´ シンプルï¼ã¨ã«ã‹ãメソッドを細ã‹ãã‚ã‘ã¦çŸã„ã‚‚ã®ã«ã™ã‚‹çš„ãªæ„味ã«ã¨ã‚Œã‚‹ã¨ã„ã†æŒ‡æ‘˜ã‚’å—ã‘ã¾ã—ãŸã®ã§è£œè¶³ã‚’追記 並列処ç†ã¨ä¸¦è¡Œå‡¦ç†ã®é•
1
ランã‚ング
ランã‚ング
メンテナンス
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
{{#tags}}- {{label}}
{{/tags}}