I did a quick test with a limit of 100 and it works: Caused by: net.sf.jailer.subsetting.RowLimitExceededException: The row limit (100) has been exceeded. at net.sf.jailer.subsetting.SubsettingEngine.checkRowLimit(SubsettingEngine.java:1566) Perfect. Thank you!
Thank you, I will try out "$DISTANCE" and "$IS_SUBJECT"
Today we ran into a problem when Jailer tried to collect data indefinitely. I know this is expected behavior in most cases. We ran the exractionmodel hundreds of times with different ids and it always returned a relatively small amount of records. The data connected to the id causing the problem is probably somewhat special. A situation like this is potentially dangerous for us. Jailer runs unattended and if it collects too much or even all data it might clog up the system, causing further problems....
Today we ran into a problem when Jailer tried to collect data indefinitely. I know this is expected behavior in most cases. We ran the exractionmodel hundreds of times with different ids and it always returned a relatively small amount of records. The data connected to the id causing the problem is probably somewhat special. A situation like this is potentially dangerous for us. Jailer runs unattended and if it collects too much or even all data it might clog up the system, causing further problems....
Is it recommended to use ROWID or not? What are the factors to decide for or against ROWIDs? Up to now we always used them, but today wie ran a new extraction model with about 50 tables resulting in 550 exported rows. With "use ROWID column" enabled it took over an hour. We noticed that one query with row ids took especially long. So we disabled rowids and it was incredibly fast, only a few seconds. The problem seems to be the select union. Attached explain plans with select union and rewritten without....
Is it recommended to use ROWID or not? What are the factors to decide for or against ROWIDs? Up to now we always used them, but today wie ran a new extraction model with about 50 tables resulting in 550 exported rows. With "use ROWID column" enabled it took over an hour. We noticed that one query with row ids took especiall long. So we disabled rowids and it was incredibly fast, only a few seconds. The problem seems to be the select union. Attached explain plans with select union and rewritten without....
The tests I did were always with the same full dataset (1 table with ~1 Million rows). I also tried with 250 rows per statement with no effect.. It could be that these performance issues are specific to our environment. Databases behave differently due to different configuration, system resources, network, table size, data types and so on. I will do some more investigation and maybe also try another database. Batching definitely improves performance. I will compare batching to the performance without...
I recently exported and imported 1 Million records from and to an Oracle Database and encountered a performance problem. Our Use Case We have two databases: one central database in the internal network and an external database in the DMZ. The data is maintained in the internal database and transfered daily to the external database. We use Jailer to export the data into an sql-file with upserts and then use Jailer to execute the exported scripts on the external database. When I tested the ex- and...