Facebookã¯ãæ°ãã¿ãã¤ãç´ã®å¤§è¦æ¨¡ãã¼ã¿ã«å¯¾ãã¦ãã対話çã«ã¢ãããã¯ãªåãåãããå¯è½ã«ããåæ£SQLã¨ã³ã¸ã³ãPrestoããããªã¼ãã³ã½ã¼ã¹ã§å ¬éãã¾ããã Prestoã¯Facebook社å ã§å¤§è¦æ¨¡ãã¼ã¿ã®åæã®ããã«éçºããããã§ã«å社社å 使ããã¦ãããã®ã Facebookã¯Prestoãéçºããèæ¯ã¨ãã¦ã大éã®ãã¼ã¿ãHadoop/HDFSãã¼ã¹ã§ä¿åãããã®ã®ããããæåã®MapReduceã§ã¯ãªãããªã¢ã«ã¿ã¤ã æ§ã«åªããå¦çãå¿ è¦ã«ãªã£ãããã ã¨ã次ã®ããã«èª¬æãã¦ãã¾ãã Facebookâs warehouse data is stored in a few large Hadoop/HDFS-based clusters. Hadoop MapReduce [2] and Hive are designed for large-scale, reliabl
ãªãªã¼ã¹ãé害æ å ±ãªã©ã®ãµã¼ãã¹ã®ãç¥ãã
ææ°ã®äººæ°ã¨ã³ããªã¼ã®é ä¿¡
å¦çãå®è¡ä¸ã§ã
j次ã®ããã¯ãã¼ã¯
kåã®ããã¯ãã¼ã¯
lãã¨ã§èªã
eã³ã¡ã³ãä¸è¦§ãéã
oãã¼ã¸ãéã
{{#tags}}- {{label}}
{{/tags}}