Road of big data week07 - day06 (Sqoop some problems when exporting from HDFS into a relational database)

Question one:

Encounter this problem during the upload process:

ERROR tool.ExportTool: Encountered IOException running export job: java.io.IOException: No columns to generate for ClassWriter

 

 Solution:

When searching online version of the driver is too low due to a lot of blog said that when, in fact, try this method when we can start with this: add a command line, - drsiver com.mysql.jdbc.Driver \ then the problem is solved! ! !

 

 

Question two:

Export is still time, will complain, but we are very amazing discovery, as well as some data is imported. This is the next question.

Caused by: java.lang.NumberFormatException: For input string: "null"

 

 Solution: Because the data resulting from the presence of null worth

Add a line in the command (in a manner of modifications, Second way is to change my format): - input-null-string    '\\ N' \       and some bloggers say that the increase in two rows, but I can also add a line specific reasons no scrutiny.

 

Q Leg three:

java.lang.RuntimeException: Can't parse input data: '1998/5/11'

 

 Problems like this occur, mostly because the data with relational database on HDFS create a table of field type mismatch caused. After careful comparison modified, this would not have given it! !

 

 

Of course, the problem will certainly not happen to such a point, mainly rely on the usual accumulation, modification, in order to continue to grow.

Note: Do not make like a knock on the wrong letter, command or wrong, less space to play such a stupid mistake, this will be very time-consuming, the only solution, that is, more practice, diligence knock. ! ! !

Guess you like

Origin www.cnblogs.com/wyh-study/p/12078378.html
Recommended