Greenplum table 之 外部表

轉載自：

https://www.cnblogs.com/kingle-study/p/10552097.html

一、外部表介紹

　　Greenplum 在數據加載上有一個明顯的優勢,就是支持數據的並發加載,gpfdisk是並發加載的工具,數據庫中對應的就是外部表

　　所謂外部表，就是在數據庫中只有表定義、沒有數據，數據都存放在數據庫之外的數據文件。greenplum可以對一個外部表執行正常的DML操作，當讀取數據的時候，數據庫從數據文件中加載數據。外部表支持在segment上並發地告訴從gpfdist導入數據，由於是從segment上導入數據，所以效率很高。

　　結構圖:

　　外部表需要指定gpfdist的IP和端口，還要有詳細的目錄地址，文件名支持通配符匹配。可以編寫多個gpfdist地址，但是總數不能超過總的segment數量，否則會報錯。

　　GPDB提供兩種外部表：可讀外部表用於數據裝載、可寫外部表用於數據卸載。外部表可基於文件、亦可基於WEB，這兩種都能實現可讀、可寫。

當一個查詢使用一個常規的外部表，該外部表被認為是可重讀的，因為在該查詢期間數據是靜態的。而對於WEB外部表，數據是不可重讀的，因為在該查詢的執行期間數據可能會發生變化。

　　　　可寫外部表用以從數據庫表中選擇記錄並輸出到文件、命名管道或其他可執行程序。 比如，可以從GPDB中卸載數據並發送到一個可執行程序，該程序連接到其他數據庫或者ETL工具並裝載數據到其他地方。 可寫外部表還可以用於輸出到GPDB的並行MapReduce計算。

　　　　可寫外部表被定義后，數據即可從數據庫表中被選擇並插入到該可寫外部表。 可寫外部表只允許INSERT操作 – SELECT、 UPDATE、 DELETE或TRUNCATE是不允許的。可寫外部表輸出數據到一個可執行程序，該程序要能夠接受流輸入數據。

　　　　在創建外部表的時候，可以指定分隔符、err表、指定允許出錯的數據條數，以及源文件的編碼等信息。

二、外部表語法

CREATE [READABLE] EXTERNAL TABLE table_name    
    ( column_name data_type [, ...] | LIKE other_table )
      LOCATION ('file://seghost[:port]/path/file' [, ...])
        | ('gpfdist://filehost[:port]/file_pattern[#transform]'
        | ('gpfdists://filehost[:port]/file_pattern[#transform]'
            [, ...])
        | ('gphdfs://hdfs_host[:port]/path/file')
      FORMAT 'TEXT'
            [( [HEADER]
               [DELIMITER [AS] 'delimiter' | 'OFF']
               [NULL [AS] 'null string']
               [ESCAPE [AS] 'escape' | 'OFF']
               [NEWLINE [ AS ] 'LF' | 'CR' | 'CRLF']
               [FILL MISSING FIELDS] )]
           | 'CSV'
            [( [HEADER]
               [QUOTE [AS] 'quote']
               [DELIMITER [AS] 'delimiter']
               [NULL [AS] 'null string']
               [FORCE NOT NULL column [, ...]]
               [ESCAPE [AS] 'escape']
               [NEWLINE [ AS ] 'LF' | 'CR' | 'CRLF']
               [FILL MISSING FIELDS] )]
           | 'AVRO'
           | 'PARQUET'
 
           | 'CUSTOM' (Formatter=<formatter specifications>)
     [ ENCODING 'encoding' ]
     [ [LOG ERRORS [INTO error_table]] SEGMENT REJECT LIMIT count
       [ROWS | PERCENT] ]
 
CREATE [READABLE] EXTERNAL WEB TABLE table_name    
   ( column_name data_type [, ...] | LIKE other_table )
      LOCATION ('http://webhost[:port]/path/file' [, ...])
    | EXECUTE 'command' [ON ALL
                          | MASTER
                          | number_of_segments
                          | HOST ['segment_hostname']
                          | SEGMENT segment_id ]
      FORMAT 'TEXT'
            [( [HEADER]
               [DELIMITER [AS] 'delimiter' | 'OFF']
               [NULL [AS] 'null string']
               [ESCAPE [AS] 'escape' | 'OFF']
               [NEWLINE [ AS ] 'LF' | 'CR' | 'CRLF']
               [FILL MISSING FIELDS] )]
           | 'CSV'
            [( [HEADER]
               [QUOTE [AS] 'quote']
               [DELIMITER [AS] 'delimiter']
               [NULL [AS] 'null string']
               [FORCE NOT NULL column [, ...]]
               [ESCAPE [AS] 'escape']
               [NEWLINE [ AS ] 'LF' | 'CR' | 'CRLF']
               [FILL MISSING FIELDS] )]
           | 'CUSTOM' (Formatter=<formatter specifications>)
     [ ENCODING 'encoding' ]
     [ [LOG ERRORS [INTO error_table]] SEGMENT REJECT LIMIT count
       [ROWS | PERCENT] ]
 
CREATE WRITABLE EXTERNAL TABLE table_name
    ( column_name data_type [, ...] | LIKE other_table )
     LOCATION('gpfdist://outputhost[:port]/filename[#transform]'
      | ('gpfdists://outputhost[:port]/file_pattern[#transform]'
          [, ...])
      | ('gphdfs://hdfs_host[:port]/path')
      FORMAT 'TEXT'
               [( [DELIMITER [AS] 'delimiter']
               [NULL [AS] 'null string']
               [ESCAPE [AS] 'escape' | 'OFF'] )]
          | 'CSV'
               [([QUOTE [AS] 'quote']
               [DELIMITER [AS] 'delimiter']
               [NULL [AS] 'null string']
               [FORCE QUOTE column [, ...]] ]
               [ESCAPE [AS] 'escape'] )]
           | 'AVRO'
           | 'PARQUET'
 
           | 'CUSTOM' (Formatter=<formatter specifications>)
    [ ENCODING 'write_encoding' ]
    [ DISTRIBUTED BY (column, [ ... ] ) | DISTRIBUTED RANDOMLY ]
 
CREATE WRITABLE EXTERNAL WEB TABLE table_name
    ( column_name data_type [, ...] | LIKE other_table )
    EXECUTE 'command' [ON ALL]
    FORMAT 'TEXT'
               [( [DELIMITER [AS] 'delimiter']
               [NULL [AS] 'null string']
               [ESCAPE [AS] 'escape' | 'OFF'] )]
          | 'CSV'
               [([QUOTE [AS] 'quote']
               [DELIMITER [AS] 'delimiter']
               [NULL [AS] 'null string']
               [FORCE QUOTE column [, ...]] ]
               [ESCAPE [AS] 'escape'] )]
           | 'CUSTOM' (Formatter=<formatter specifications>)
    [ ENCODING 'write_encoding' ]
    [ DISTRIBUTED BY (column, [ ... ] ) | DISTRIBUTED RANDOMLY ]

三、創建外部表

　　01,語法

gpfdist [-d directory] [-p http_port] [-l log_file] [-t timeout]
[-S] [-w time] [-v | -V] [-s] [-m max_length] [--ssl certificate_path]
gpfdist -? | --help
gpfdist --version

　　02,啟動進程

--創建gpdist進程
[gpadmin@greenplum02 ~]$ mkdir script
[gpadmin@greenplum02 ~]$ nohup gpfdist -d /home/gpadmin/script/ -p 8081 -l /home/gpadmin/script/gpfdist.log &
[1] 6904
[gpadmin@greenplum02 ~]$ nohup: ignoring input and appending output to ‘nohup.out’
[gpadmin@greenplum02 ~]$ ss -lntup|grep 8081
tcp    LISTEN     0      128      :::8081                 :::*                   users:(("gpfdist",pid=6904,fd=6))

---配置讀取文件
[gpadmin@greenplum02 script]$ cat test.txt
Prague,Jan,101,4875.33
Rome,Mar,87,1557.39
Bangalore,May,317,8936.99
Beijing,Jul,411,11600.67
San Francisco,Sept,156,6846.34
Paris,Nov,159,7134.56
San Francisco,Jan,113,5397.89
Prague,Dec,333,9894.77
Bangalore,Jul,271,8320.55
Beijing,Dec,100,4248.41

q
[gpadmin@greenplum02 script]$ pwd
/home/gpadmin/script
--后面的是錯誤信息

　　03,創建外部表

create external table public.test
(
country varchar(128),
name varchar(128),
id int,
sale varchar(128)
)
location ('gpfdist://192.168.0.222:8081/test.txt')
format 'text'
(delimiter ',' null as '' escape 'off')
encoding 'utf8'
log error segment reject limit 3 rows;

--- location   文件所在位置，可以直接是本地路徑、gpfdist地址、gpfdists地址、gphdfs地址。
--- format     文本類型
--- delimiter  分隔符
--- encoding    編碼
--- log error into  錯誤數據表，記錄錯誤數據，會自動創建。一般都是tablename_err格式，例如t1_err。
--- segment reject limit  錯誤數據的條數/百分比(rows/percent)，超過設置值會報錯。最小值是2。用來確保數據的完整性。

結果:
postgres=# create external table public.test99(country varchar(128),name varchar(128),id int,sale varchar(128))location ('gpfdist://192.168.0.222:8081/test.txt')format 'text'(delimiter ',' null as '' escape 'off')encoding 'utf8'log errors segment reject limit 3 rows;
CREATE EXTERNAL TABLE
postgres=# SELECT * from public.test99
postgres-# ;
NOTICE:  Found 2 data formatting errors (2 or more input rows). Rejected related input data.
    country    | name | id  |   sale
---------------+------+-----+----------
 Prague        | Jan  | 101 | 4875.33
 Rome          | Mar  |  87 | 1557.39
 Bangalore     | May  | 317 | 8936.99
 Beijing       | Jul  | 411 | 11600.67
 San Francisco | Sept | 156 | 6846.34
 Paris         | Nov  | 159 | 7134.56
 San Francisco | Jan  | 113 | 5397.89
 Prague        | Dec  | 333 | 9894.77
 Bangalore     | Jul  | 271 | 8320.55
 Beijing       | Dec  | 100 | 4248.41
(10 rows)

postgres=# SELECT * from test99;
NOTICE:  Found 2 data formatting errors (2 or more input rows). Rejected related input data.
    country    | name | id  |   sale
---------------+------+-----+----------
 Prague        | Jan  | 101 | 4875.33
 Rome          | Mar  |  87 | 1557.39
 Bangalore     | May  | 317 | 8936.99
 Beijing       | Jul  | 411 | 11600.67
 San Francisco | Sept | 156 | 6846.34
 Paris         | Nov  | 159 | 7134.56
 San Francisco | Jan  | 113 | 5397.89
 Prague        | Dec  | 333 | 9894.77
 Bangalore     | Jul  | 271 | 8320.55
 Beijing       | Dec  | 100 | 4248.41
(10 rows)

　　04,數據裝載

insert into table select * from table_ext;

內部表<----外部表