Creating the dumpfile
This is a sample for MySql.
1.Stop the JobScheduler
2.Create a dump of history tables
./mysqldump scheduler SCHEDULER_HISTORY SCHEDULER_ORDER_HISTORY SCHEDULER_ORDER_STEP_HISTORY SCHEDULER_VARIABLES --host=<host> --port=<port> --user=<user> -p --add-drop-table --skip-comments --single-transaction --opt > scheduler_history.dmp
3.create an archive database
prompt:~> /opt/mysql-5.0.15/bin/mysql --host=<host> --port=<port> --user=<user> -p scheduler_archive mysql> create database scheduler_archive mysql> exit
4.import dump to the archive
prompt:~>./mysql --host=<host> --port=<port> --user=<user> -p scheduler_archive < scheduler_history.dmp
5.Optional drop the archived tables (DO NOT DROP SCHEDULER_VARIABLES!!!!!!!!!!!!!!!!!)
prompt:~> ./mysql --host=<host> --port=<port> --user=<user> -p scheduler mysql> drop table SCHEDULER_HISTORY mysql> drop table SCHEDULER_ORDER_HISTORY mysql> drop table SCHEDULER_ORDER_STEP_HISTORY mysql> exit
6.Recreate the tables
prompt:~> ./opt/sos-berlin/..../install/scheduler_install_tables.sh
How to access logs from the archive by id
Setup a new job scheduler instance. This instance should have no own jobs.
To access a task log
http://localhost:4444/show_log?task=1262811
To access an order log
http://localhost:4444/show_log?order=backup_pg_sql&history_id=403816