oracle高級(jí)技術(shù)之大數(shù)據(jù)處理-文庫(kù)吧資料
2025-02-20 15:01本頁(yè)面
【正文】 的更名（ RENAME））? 分區(qū)索引的分離分區(qū)索引的分離 （（ SPLIT））? 分區(qū)索引的分區(qū)索引的 Unusable分區(qū)表和 Local索引OCT2023SEP2023AUG2023NOV 2023OCT2023新月份數(shù)據(jù)的加載和索引的維護(hù)NOV2023NOV2023NOV2023NOV2023NOV2023NOV2023NOV2023NOV2023...“滾動(dòng)窗口 ”操作 大量數(shù)據(jù)高速裝載OCT2023SEP2023NOV2023 NOV 2023 NOV 2023 OCT2023OCT2023SEP2023NOV2023 DEC 2023 NOV 2023 OCT2023OCT2023SEP2023NOV2023 DEC 2023 NOV 2023 OCT2023OCT2023SEP2023NOV2023 DEC 2023 NOV 2023 OCT2023OCT2023SEP2023NOV2023 DEC 2023 NOV 2023 OCT2023OCT2023SEP2023NOV2023 DEC 2023 NOV 2023 OCT2023OCT SEP2023NOV2023刪除或歸檔最老月份的數(shù)據(jù)OCT2023新月份數(shù)據(jù)的加載和索引的維護(hù)...分區(qū)表和 Local索引“滾動(dòng)窗口 ”操作 大量數(shù)據(jù)高速裝載分區(qū)交換功能? 通過(guò)交換數(shù)據(jù)段，實(shí)現(xiàn)分區(qū)和非分區(qū)表的數(shù)據(jù)交換。 ? 并行數(shù)據(jù)操作：對(duì)于經(jīng)常執(zhí)行并行操作（如 Parallel Insert,Parallel Update等）的表應(yīng)考慮進(jìn)行分區(qū)。? 數(shù)據(jù)備份和恢復(fù)： 按時(shí)間周期進(jìn)行表空間的備份時(shí)，將分區(qū)與表空間建立對(duì)應(yīng)關(guān)系。 ? 數(shù)據(jù)維護(hù)：按時(shí)間段刪除成批的數(shù)據(jù)，例如按月刪除歷史數(shù)據(jù)。 ? 數(shù)據(jù)訪問(wèn)特性：基于表的大部分查詢應(yīng)用，只訪問(wèn)表中少量的數(shù)據(jù)。ddmonyyyy39。28jun202339。ddmonyyyy39。25jun202339。partitionprocessslavesparallelkey.? Oracletheisandhashifpartitionatablehashjoiningoccurkey.? Partitionwisethepartitionedtablesofonlyoccurspartitionwisepartitions.? Apairsjoinsjoinsplitsjoinonaretablestwowhenjoinfullequipartitioned.identicallyareindexes) Partition Partition Operation Options Object Name Start Stop SELECT STATEMENT PARTITION RANGE INLIST KEY(INLIST) KEY(INLIST) TABLE ACCESS FULL DAILY_TRANS_SUMM KEY(INLIST) KEY(INLIST) Nonpartitionwise joinFull partitionwise joinPartial partitionwise joinQuery slave Partition Partitioned table1 23PartitionWise JoinPartitionwise Joins? Tables,39。) Partition Partition Operation Options Object Name Start Stop SELECT STATEMENT PARTITION RANGE ITERATOR 231 234 TABLE ACCESS FULL DAILY_TRANS_SUMM 231 234 1 select * from daily_trans_summ 2* where trans_datetime in (39。,39。) and to_date(39。,39。分區(qū)裁剪舉例1 select * from daily_trans_summ 2* where trans_datetime between to_date(39。local。onindextablespacecust_index07,partitioncust_idx_p07tablespacecust_index05,partitioncust_idx_p05tablespacecust_index03,partitioncust_idx_p03tablespacecust_index01,partitioncust_idx_p01customers(customer_no)cust_idx2cust_index04)create(MAXVALUE)lesscust_p04tablespacethanvaluescust_index02,partition(‘N’)lesscust_p02tablespacethanvalues(customer_name)(partitionbycustomers(customer_name)globalcust_idx1columns.? 分區(qū)表上的非分區(qū)索引等同于 Global索引分區(qū)索引? Global索引必須是范圍分區(qū) 9i之前? Global索引可以是 HASH分區(qū) 10g新特性? Global索引不支持 Bitmap索引? Unique索引必須是 prefixed， 或者包含分區(qū)字段? Local索引（ nonprefixed, nonunique） 可以不包含分區(qū)字段createtheprefixapartitioneditprefixedindexindexes？分區(qū)索引不分區(qū) 分區(qū)不分區(qū)local)) subpartitions 4 )復(fù)合分區(qū)圖示復(fù)合分區(qū)特點(diǎn)? Oracle支持的 Composite分區(qū)：? RangeHash,RangeList? 既適合于歷史數(shù)據(jù)，又適合于數(shù)據(jù)均勻分布? 與范圍分區(qū)一樣提供高可用性和管理性? 更好的 PDML和 partitionwise,39。)) ,partition dtd_20230622 values less than (to_date(39。,39。)) (subpartition dtd_20230620_s01 ,subpartition dtd_20230620_s02 ,subpartition dtd_20230620_s03 tablespace dtd_data03 ,subpartition dtd_20230620_s04 tablespace dtd_data04 ,subpartition dtd_20230620_s05 tablespace dtd_dat
公司管理相關(guān)推薦
鄂ICP備17016276號(hào)-1