Summary
This issue raises from time to time, and here we focus on one of the causes. What is about writing log to disk.
Error
What are the errors to check:
You will need to check if in the run log file, you have following lines:
o WARN : Move data from oltp tables to history tables thread 6: root: Exception in Task java.lang.StackOverflowError
o at com.outerbay.foundation.services. dbms.GenericDAO.logSqlError(GenericDAO.java:1251)
Cause
This applies for SDM versions.
What is the problem?
Customer launches BF and this almost finish, but fails at the end, however, in WC Monitoring we can see all main stages talking to selection/Copy/Purge were done.
Process try to move memory log lines about featuare running like BF, scanning, to disk.
Trying to write it down downloading it from memory to disk, but, disk is full, then process last steps get fail.
Fix
- There are many fails in BFs showing memory stack overflow, so, here, this solution is not the only one.
- However, first think to check is about disk space where the log file is being stored.
- The issue is because BF found ‘disk full’ trying to write memory-log-lines down to disk, so, no room to do that task.
- Please check this out with customer and recover the BF.
- Analyze results.