蜂巢失敗：ParseException的線路2：0無法識別「」附近MACADDRESS「」「CHAR」「（」輸入列規範

Question

我試着運行hive -v -f sqlfile.sql

下面是該文件的內容

CREATE TABLE UpStreamParam (
'macaddress' CHAR(50), 
'datats' BIGINT, 
'cmtstimestamp' BIGINT, 
'modulation' INT, 
'chnlidx' INT, 
'severity' BIGINT, 
'rxpower' FLOAT, 
'sigqnoise' FLOAT, 
'noisedeviation' FLOAT, 
'prefecber' FLOAT, 
'postfecber' FLOAT, 
'txpower' FLOAT, 
'txpowerdrop' FLOAT, 
'nmter' FLOAT, 
'premtter' FLOAT, 
'postmtter' FLOAT, 
'unerroreds' BIGINT, 
'corrected' BIGINT, 
'uncorrectables' BIGINT) 
STORED AS ORC TBLPROPERTIES ("orc.compress"="SNAPPY","orc.bloom.filters.columns"="macaddress") 
PARTITIONED BY ('cmtsid' CHAR(50),' date' INT) 
LOCATION '/usr/hive/warehouse/UpStreamParam' ; 

而且我得到以下錯誤：

FAILED: ParseException line 2:0 cannot recognize input near ''macaddress'' 'CHAR' '(' in column specification

Answer 1

第一，列名必須由`（反引號）包圍，不'（單引號）

因此，您必須將'macaddress'更換爲`macaddress`以及所有其他列名稱。

二，訂單號爲STORED AS和TBLPROPERTIES和PARTITIONED BY和LOCATION是錯誤的。 正確的順序是PARTITIONED BY,STORED AS,LOCATION,TBLPROPERTIES。

有關詳細信息，請參閱蜂巢語言手冊。 https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDL#LanguageManualDDL-CreateTable

所以正確的代碼是

CREATE TABLE UpStreamParam (
`macaddress` CHAR(50), 
`datats` BIGINT, 
`cmtstimestamp` BIGINT, 
`modulation` INT, 
`chnlidx` INT, 
`severity` BIGINT, 
`rxpower` FLOAT, 
`sigqnoise` FLOAT, 
`noisedeviation` FLOAT, 
`prefecber` FLOAT, 
`postfecber` FLOAT, 
`txpower` FLOAT, 
`txpowerdrop` FLOAT, 
`nmter` FLOAT, 
`premtter` FLOAT, 
`postmtter` FLOAT, 
`unerroreds` BIGINT, 
`corrected` BIGINT, 
`uncorrectables` BIGINT) 
PARTITIONED BY (`cmtsid` CHAR(50), `date` INT) 
STORED AS ORC 
LOCATION '/usr/hive/warehouse/UpStreamParam' 
TBLPROPERTIES ("orc.compress"="SNAPPY","orc.bloom.filters.columns"="macaddress"); 

Answer 2

的問題可能是因爲蜂房版本，其中CHAR數據類型從蜂巢版本0.13支持。

如果您使用的是較低版本，請嘗試使用string/varchar數據類型。

請參考， https://cwiki.apache.org/confluence/display/Hive/LanguageManual+Types#LanguageManualTypes-Char