| ²é¿´: 463 | »Ø¸´: 0 | |||
xgxofdreamÒø³æ (ÖøÃûдÊÖ)
|
[½»Á÷]
¡¾ÇóÖú¡¿Çó·ÒëÒ»¶ÎÓйØÅúÁ¿¼ÓÔØÊý¾ÝµÄÓ¢ÎÄ
|
|
·ÒëһƪרÀû£¬Éæ¼°µ½Í¦¶àÇ°ÑØµÄ¸ÅÄ¹ý²»È¥¡£Çë´ó¼Ò°ïæÁË~£¡ The updating of large databases with large amounts of new information presents challenges. In particular, the records in large database may be stored in aplurality of partitions with one or more partitions being handled by a server that may form part of a server system. Where a server stores multiple partitions, the partitions may or may not contain adta in which the key ranges are sequential. For example, a server may store a first partition with one key range,and may also store a second partition with a different key range.the key range of such second partition might not start where the key range of the first partition leaves off,but rather may be a range that is much higher or much lower than the key range of the first partition. Bulk loading of data into the database may be accomplished by inserting new data into the appropriate partition handled by the appropriate server. As a partition reaches its maximum desired size, the partitions may be divided, or split, into two parts. After completion of the bulk loading process, the partitions may be balanced among the various servers of the server system, for example, so that the partitions are more evenly distributed across the servers on which they are stored. [ Last edited by xgxofdream on 2010-4-29 at 00:40 ] |
» ²ÂÄãϲ»¶
293µ÷¼Á
ÒѾÓÐ3È˻ظ´
»úе¹¤³Ì264ѧ˶Çóµ÷¼Á
ÒѾÓÐ3È˻ظ´
264Çóµ÷¼Á
ÒѾÓÐ9È˻ظ´
Çóµ÷¼Á
ÒѾÓÐ6È˻ظ´
22408 266Çóµ÷¼Á
ÒѾÓÐ7È˻ظ´
Çóµ÷¼Á
ÒѾÓÐ14È˻ظ´
273Çóµ÷¼Á
ÒѾÓÐ43È˻ظ´
307Çóµ÷¼Á
ÒѾÓÐ13È˻ظ´
327Çóµ÷¼Á
ÒѾÓÐ7È˻ظ´
338Çóµ÷¼Á
ÒѾÓÐ6È˻ظ´















»Ø¸´´ËÂ¥