Using an id of a table for sql_last_value in logstash?

问题

I'm having a MySQL statement as such within my jdbc plugin in logstash input.

statement => "SELECT * from TEST where id > :sql_last_value"

My table doesn't have any date or datetime field as such. So I'm trying to update the index, by checking minute by minute using a scheduler, whether any new rows have been added to the table.

I should only be able to update the new records, rather than updating the existing value changes from an existing record. So to do this I'm having this kinda of a logstash input:

input {
    jdbc {
        jdbc_connection_string => "jdbc:mysql://myhostmachine:3306/mydb" 
        jdbc_user => "root"
        jdbc_password => "root"
        jdbc_validate_connection => true
        jdbc_driver_library => "/mypath/mysql-connector-java-5.1.39-bin.jar"
        jdbc_driver_class => "com.mysql.jdbc.Driver"
        jdbc_paging_enabled => "true"
        jdbc_page_size => "50000"
        schedule => "* * * * *"
        statement => "SELECT * from mytable where id > :sql_last_value"
        use_column_value => true
        tracking_column => id
        last_run_metadata_path => "/path/.logstash_jdbc_last_run"
        clean_run => true
    }
}

So whenever I create an index and run this logstash file in order to upload the docs, it doesn't get uploaded at all. The docs count shows as zero. I made sure that I deleted the .logstash_jdbc_last_run before I ran the logstash conf file.

Part of logstash console output:

[2016-11-02T16:33:00,294][INFO ][logstash.inputs.jdbc ] (0.002000s) SELECT count(*) AS count FROM (SELECT * from TEST where id > '2016-11-02 11:02:00') AS t1 LIMIT 1

and this keeps on going by checking minute by minute which is correct but then it doesn't get the records. How does it work?

Am I missing something? Any help could be appreciated.

回答1:

You need to modify your logstash configuration like this:

jdbc { 
  jdbc_connection_string => "jdbc:mysql://myhostmachine:3306/mydb" 
  jdbc_user => "root" 
  jdbc_password => "root" 
  jdbc_validate_connection => true 
  jdbc_driver_library => "/mypath/mysql-connector-java-5.1.39-bin.jar" 
  jdbc_driver_class => "com.mysql.jdbc.Driver" 
  jdbc_paging_enabled => "true" 
  jdbc_page_size => "50000" 
  schedule => "* * * * *" 
  statement => "SELECT * from TEST where id > :sql_last_value" 
  use_column_value => true 
  tracking_column => "id" 
  tracking_column_type => "numeric" 
  clean_run => true 
  last_run_metadata_path => "/mypath/.logstash_jdbc_last_run" 
}

The last five settings are important in your case. Also make sure to delete the .logstash_jdbc_last_run file even though clean_run => true does it.

来源：https://stackoverflow.com/questions/40378580/using-an-id-of-a-table-for-sql-last-value-in-logstash