We are using LogStash version 7.3.2 to fetch SQL Server data.
And it is working fine but sometimes it is throwing below exception:
Exception when executing JDBC query {:exception=>#
transition (daylight savings time 'gap'): 1942-09-01T00:00:00.000 (Asia/Kolkata)>}
When I check in SQL server then there is no value like 1942-09-01T00:00:00.000.
My LogStash config is as below:
jdbc_connection_string => "jdbc:sqlserver://HOST:PORT;databaseName=DB_NAME;integratedSecurity=false
jdbc_user => "USERNAME"
jdbc_password => "PASSWORD"
jdbc_driver_library => "/home/user/LOGSTASH/mssql-jdbc-7.4.1.jre8.jar"
jdbc_driver_class => "com.microsoft.sqlserver.jdbc.SQLServerDriver"
statement => "SELECT * FROM TABLE_NAME where INCR_COLUMN > :sql_last_value "
schedule => "*/60 * * * * *"
use_column_value => true
tracking_column => "INCR_COLUMN"
tracking_column_type => "timestamp"
record_last_run => true
last_run_metadata_path => "/home/user/LOGSTASH/LAST_RUN/.mssql_USERS_logstash_jdbc_last_run"
connection_retry_attempts => "1000"
connection_retry_attempts_wait_time => "60"
jdbc_default_timezone => "UTC"
Note: In SQL Server INCR_COLUMN is defined as DateTime.
Any help regarding this would be highly appreciated.
In Logstash if a
datetime
string has noTZ part
it is assumed to be aUTC time
.If you know that the datetime is a local time string then you need to tell the date filter which timezone the date was generated in. You may use one of the Joda timezones,
e.g.
America/New_York
,America/Detroit
orEST5EDT
- these are merely labels that point to Timezone Java code that know what changes in clocks occurred in all timezones for quite some time into the past.See this page for info on how the updates are followed. http://www.oracle.com/technetwork/java/javase/tzdata-versions-138805.html 100
Once the datetime string is converted to a Timestamp object (by the date filter) it is considered
UTC time
.