[B! DBMS] peketaminã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

peketamin id:peketamin

DBMSã«é–¢ã™ã‚‹peketaminã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (7)

${{author_name}}$

{{{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}}

æƒ…å ±ã‚·ã‚¹ãƒ†ãƒ ç‰¹åˆ¥è¬›ç¾©D #GB37301
GB37301.md æƒ…å ±ã‚·ã‚¹ãƒ†ãƒ ç‰¹åˆ¥è¬›ç¾©D 1æ™‚é™ç›® ã‚¤ãƒ³ãƒˆãƒ - å·å³¶å…ˆç”Ÿ(ç‘æ³¢å¤§å¦) 2æ™‚é™ç›® Inside PostgreSQL Kernel - æ°¸å®‰ æ‚Ÿå²ã•ã‚“(ã‚¢ãƒƒãƒ—ã‚¿ã‚¤ãƒ ãƒ»ãƒ†ã‚¯ãƒŽãƒã‚¸ãƒ¼ã‚º) 3æ™‚é™ç›® ãƒ‡ãƒ¼ã‚¿ã‚¹ãƒˆãƒ¬ãƒ¼ã‚¸ã®è«¸ã€… - æ˜Ÿé‡Ž å–¬ã•ã‚“(ã‚µã‚¤ãƒœã‚¦ã‚ºãƒ»ãƒ©ãƒœ) 4æ™‚é™ç›® ä¸¦åˆ—ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚·ã‚¹ãƒ†ãƒ ã®æ¦‚å¿µã¨åŽŸç† - æ²¹äº• èª ã•ã‚“(ç”£ç·ç ”) 5æ™‚é™ç›® Googleã‚¯ãƒ©ã‚¦ãƒ‰ãŒå®Ÿç¾ã™ã‚‹å¤§è¦æ¨¡ä¸¦åˆ—ã‚¯ã‚¨ãƒªã‚µãƒ¼ãƒ“ã‚¹ - ä½è—¤ ä¸€æ†²ã•ã‚“(Google) 6æ™‚é™ç›® åˆ†æ•£å‡¦ç†åŸºæ¿Hadoop/MapReduce - å°æ²¢ å¥å²ã•ã‚“(NTTç ”) 7æ™‚é™ç›® Treasure Data Techno logies - ä¸å· çœŸå®ã•ã‚“(Treasure Data) 8æ™‚é™ç›® Retrospection/Prospection / Norikra in Action - ç”°ç± è¡ã•ã‚“(LINEæ ªå¼ä¼šç¤¾) 9æ™‚
peketamin 2014/02/01
DBMS

SQL

åˆ†æ•£

ä¸¦åˆ—
ãƒªãƒ³ã‚¯
The Traditional RDBMS Wisdom is All Wrong
The Traditional RDBMS Wisdom is All Wrong by Michael Stonebraker Traditional RDBMS Wisdom ïµData is in disk block formatting (heavily encoded) ïµWith a main memory buffer pool of blocks ïµQuery plans ïµOptimize CPU, I/O ïµFundamental operation is read a row ïµIndexing via B-trees ïµClustered or unclustered Traditional RDBMS Wisdom ïµDynamic row-level locking ïµAries-style write-ahead log ïµReplication (asyn
peketamin 2013/10/18
RDBMS

ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹

DBMS
ãƒªãƒ³ã‚¯
SQLå…¥é–€æ›¸ã¨ã—ã¦ã€ŒH2å¯¾å¿œã€ã—ãŸã‹ã£ãŸç†ç”± - flairDays - ã¦ã•ãã‚Šã®æ—¥ã€…
ä¹…ã—ã¶ã‚Šã«ã€Œã‚¿ãƒƒãƒã€ã‚’è¦‹ã¦ãƒ‰ã‚ãƒ‰ã‚ã—ã¾ã—ãŸ GWå‰åŠ2æ—¥ç›®ã€çš†æ§˜ã„ã‹ãŒãŠéŽã”ã—ã§ã—ã‚‡ã†ã‹ã€‚ ã€Œå¼•ãã“ã‚‚ã‚Šç³»ã‚¨ãƒ³ã‚¸ãƒ‹ã‚¢ã€ã‚’è‡ªç§°ã™ã‚‹ç§ã¯ã€ã›ã£ã‹ãå¤–å‡ºã—ãŸã®ã«ã™ãå¸°ã£ã¦ãã¦ã—ã¾ã„ã€PCã¨ãƒ†ãƒ¬ãƒ“ã®å‰ã§å¹³æ—¥ã¨å¤‰ã‚ã‚‰ãªã„ãŠä»•äº‹ãƒ©ã‚¤ãƒ•ã§ã™ã€‚ ãƒ†ãƒ¬ãƒ“ã¨è¨€ãˆã°å…ˆæ—¥ã€ã€Œã‚¿ãƒƒãƒã€ãŒå†æ”¾é€ã—ã¦ã„ã¾ã—ãŸã€‚ã¿ãªã¿ã¡ã‚ƒã‚“ã€ã„ã¤è¦‹ã¦ã‚‚å¯æ„›ã„ã‚ã...*1ã€‚ãã—ã¦å¦„æƒ³ã‚’ã‹ããŸã¦ã¾ãã‚‹ä¸»é¡ŒæŒã«èŒãˆæ®ºã•ã‚Œã¾ãã£ãŸã‚ã®æ—¥ã€…ã€‚ ã¨ã„ã†ã“ã¨ã§*2ã€ä»Šå›žã¯ã€å…ˆæ—¥ã®ãƒ–ãƒã‚°ã§ç´¹ä»‹ã—ãŸã€Žã‚¹ãƒƒã‚ãƒªã‚ã‹ã‚‹ SQL å…¥é–€ã€ã§å…¬å¼ã«å¯¾å¿œã‚’ã†ãŸã£ãŸã€æˆ‘ã‚‰é–‹ç™ºè€…ã®å¼·ã„å‘³æ–¹ã€ŒH2 Databaseã€ã‚’ç´¹ä»‹ã—ã¦ã„ã“ã†ã¨æ€ã„ã¾ã™ã€‚ H2 Databaseã¨ã¯ H2 Databaseã¯ã€Javaã ã‘ã§è¨˜è¿°ã•ã‚Œã¦ã„ã‚‹ã‚ªãƒ¼ãƒ—ãƒ³ã‚½ãƒ¼ã‚¹ã®RDBMSã§ã™ã€‚ã‚‚ã¨ã‚‚ã¨HSQL DBã‚’ä½œã£ã¦ã„ãŸãƒˆãƒ¼ãƒžã‚¹ãƒ»ãƒŸãƒ¥ãƒ¼ãƒ©ãƒ¼ã•ã‚“*3ãŒã€ã‚µãƒ©ã‹ã‚‰ä½œã‚Šç›´ã—ãŸã‚‚ã®ã§ã™ã€‚ ã€Œå‹•ä½œãŒè¶…é€Ÿã„ã€ã¨ã„
peketamin 2013/04/29
DBMS

java

ãƒ†ã‚¹ãƒˆ
ãƒªãƒ³ã‚¯
ã‚¤ãƒãƒ¼ãƒ–ãƒ©åž‹NoSQLã®ã¾ã¨ã‚ï¼ˆMemcachedã€Redisã€Scalarisã€Tokyo Cabinet/Tyrantç·¨ï¼‰
åº ç« ãƒ“ãƒƒã‚°ãƒ‡ãƒ¼ã‚¿ã®æ™‚ä»£ ç¬¬1ç« NOSQLã¨ã¯ä½•ã‹? ç¬¬2ç« NOSQLã®ãƒ‡ãƒ¼ã‚¿ãƒ¢ãƒ‡ãƒ« ç¬¬3ç« ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã®åŸºæœ¬æ¦‚å¿µã¨æŠ€è¡“ ç¬¬4ç« Hadoopã¯NOSQL? ç¬¬5ç« ä¸»ãªNOSQLãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹è£½å“ ç¬¬6ç« NOSQLãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®é¸æŠžåŸºæº– ç¬¬7ç« NOSQLã‚’ä½¿ã†ãƒ“ã‚¸ãƒã‚¹ æœ¬é€£è¼‰ã¯æ›¸ç±ã€ŽNOSQLã®åŸºç¤ŽçŸ¥è˜ã€ï¼ˆãƒªãƒƒã‚¯ãƒ†ãƒ¬ã‚³ãƒ åˆŠã€ISBNï¼š978-4897978871ï¼‰ã§è§£èª¬ã•ã‚Œã¦ã„ã‚‹å†…å®¹ã‹ã‚‰ä¸€éƒ¨ã‚’æŠœç²‹ã—ã€æœ¬é€£è¼‰å‘ã‘ã«ä¸€éƒ¨å†ç·¨é›†ã—ã¦æŽ²è¼‰ã—ãŸã‚‚ã®ã§ã™ã€‚ æ›¸ç±ã§ã¯ã€ä¸€èˆ¬ã«NoSQLã¨å‘¼ã°ã‚Œã¦ã„ã‚‹å„ç¨®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹æŠ€è¡“ã«ã¤ã„ã¦ã€åŸºæœ¬æ¦‚å¿µã‹ã‚‰ä¸»è¦ãªãƒ—ãƒãƒ€ã‚¯ãƒˆã®ç‰¹æ€§ã€ãƒ™ãƒ³ãƒãƒžãƒ¼ã‚¯çµæžœã¾ã§ã‚’ç´¹ä»‹ã—ã¦ã„ã¾ã™ã€‚ãƒ‡ãƒ¼ã‚¿ãƒ¢ãƒ‡ãƒ«ã‚„ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã®é•ã„ã¨ã„ã£ãŸåŸºæœ¬æ¦‚å¿µã‹ã‚‰ã€å„ãƒ—ãƒãƒ€ã‚¯ãƒˆã®ç‰¹å¾´ã‚’ç†è§£ã§ãã‚‹å†…å®¹ã«ãªã£ã¦ã„ã¾ã™ã€‚ æœ¬é€£è¼‰ã§ã¯ã€ã“ã®æ›¸ç±ã®å†…å®¹ã‹ã‚‰ã€ä¸»è¦ãƒ—ãƒãƒ€ã‚¯ãƒˆã‚’ç´¹ä»‹ã—ã¦ã„ã‚‹ç¬¬5ç« ã‚’æŠœç²‹ã—ã€ãã®ã‚¨ãƒƒ
peketamin 2013/01/07
nosql

kvs

memcached

DBMS

redis
ãƒªãƒ³ã‚¯
ç¬¬6å›žã€€ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹ã¯ã©ã†ãªã®ï¼Ÿ | gihyo.jp
ã“ã®è¡¨ã§ã¯è¨€åŠã—ã¦ã„ã¾ã›ã‚“ãŒã€Consistent Hashingã‚„shardingã«ã‚ˆã£ã¦åˆ†æ•£ã•ã›ã‚„ã™ã„ã¨ã„ã†ã®ã¯ã™ã¹ã¦ã®NoSQLãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãŒå‚™ãˆã¦ã„ã‚‹ç‰¹å¾´ã§ã—ãŸãã€‚ æ¤œè¨¼ç”¨ã‚µãƒ¼ãƒã®æº–å‚™ ã§ã¯å®Ÿéš›ã®ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹ã‚’æ¯”è¼ƒã—ã¦ã¿ã¾ã—ã‚‡ã†ã€‚æ¯”è¼ƒã®ãŸã‚ã€10,000ä»¶ã®ãƒ‡ãƒ¼ã‚¿ã®æ›¸è¾¼ã¿/èªè¾¼ã¿ã‚’è¡Œã„ã¾ã™ã€‚å®Ÿè¡Œã«ã‚ãŸã‚Šå„ç¨®ã‚µãƒ¼ãƒã‚’èµ·å‹•ã—ã¾ã™ã€‚ ãƒªã‚¹ãƒˆ1ã€€æ¤œè¨¼ç”¨ã‚µãƒ¼ãƒã®èµ·å‹•ã‚³ãƒžãƒ³ãƒ‰ # memcached memcached -u nobody -d # TokyoTyrant sudo ttserver -pid /var/ttserver/pid -dmn /var/ttserver/bench.tch # Mongo DB mongod --dbpath ~/tmp/mongo db & # HBase sudo /etc/init.d/hadoop-0.20-namenode start su
peketamin 2010/12/24
ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹

nosql

ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹

mongodb

HBase

Hadoop

mysql

DBMS

KVS

db
ãƒªãƒ³ã‚¯
ã¨ã—ã®ä»Šæ—¥ã§ã
peketamin 2010/12/10
mapreduce

nosql

å¤§é‡ãƒ‡ãƒ¼ã‚¿

ä¸¦åˆ—å‡¦ç†

åˆ†æ•£å‡¦ç†

DB

DBMS

RDBMS

ä¸–ä»£
ãƒªãƒ³ã‚¯
ã‚°ãƒªãƒ¼ã®å¤§è¦æ¨¡åˆ†æ•£ã‚¹ãƒˆãƒ¬ãƒ¼ã‚¸æˆ¦ç•¥(nanofs) Vol.2 | GREE Engineering
ã¯ã˜ã‚ã« ã‚°ãƒªãƒ¼æ ªå¼ä¼šç¤¾ã§ã‚¨ãƒ³ã‚¸ãƒ‹ã‚¢ã‚’ã—ã¦ãŠã‚Šã¾ã™kgwsã¨ç”³ã—ã¾ã™ã€‚ ä»Šå›žã¯ã€å‰å›žã«å¼•ãç¶šãåˆ†æ•£ã‚¹ãƒˆãƒ¬ãƒ¼ã‚¸(nanofs)ã®HTTPãƒ¡ã‚½ãƒƒãƒ‰æ¯Žã®å‡¦ç†ã‚’ç´¹ä»‹ã•ã›ã¦ã„ãŸã ã‘ã‚Œã°ã¨æ€ã„ã¾ã™ã€‚ nanofsã¯5ã¤ã®HTTPãƒ¡ã‚½ãƒƒãƒ‰(GETã€PUTã€DELETEã€HEADã€MKCOL)ã‚’ã‚µãƒãƒ¼ãƒˆã—ã¦ãŠã‚Šã¾ã™ã€‚ä»Šå›žã¯ä¸»ãªGETã€PUTã€DELETEã®3ã¤ã«ã¤ã„ã¦ã”èª¬æ˜Žã•ã›ã¦ã„ãŸã ãã¾ã™ã€‚ ã¾ãšã¯æ§‹æˆã®ãŠã•ã‚‰ã„ nanofsã¯ã€ä¸»ã«3ã¤ã®ãƒ—ãƒã‚»ã‚¹ã§æ§‹æˆã•ã‚Œã¦ãŠã‚Šã¾ã™ã€‚ nanofsd(dispatcher) ã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã‚µãƒ¼ãƒã‹ã‚‰ãƒªã‚¯ã‚¨ã‚¹ãƒˆã‚’å—ã‘å–ã‚Šå®Ÿéš›ã«ä¿å˜ã•ã‚Œã¦ã„ã‚‹nanofsnã«æŒ¯ã‚Šåˆ†ã‘ã‚‹ 5ã¤ã®HTTPãƒ¡ã‚½ãƒƒãƒ‰ã‚’ã‚µãƒãƒ¼ãƒˆã—ã¦ã„ã‚‹(GETã€PUTã€DELETEã€HEADã€MKCOL) ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ï¼ˆKVSï¼‰ã«ä¿å˜ã—ãŸãƒ‡ãƒ¼ã‚¿ã®æƒ…å ±ã‚’é€ã‚‹ queueã«å‡¦ç†ã®æŒ‡ç¤ºã‚’é€ã‚‹ nanofsw(worke
peketamin 2010/10/13
HTTP

ã‚µãƒ¼ãƒ

åˆ†æ•£

ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹

ã‚¹ãƒˆãƒ¬ãƒ¼ã‚¸

gree

nanofs

storage

DBMS
ãƒªãƒ³ã‚¯
1