Oracle Communications EAGLE (Software)
 

PLA-Support>Sun Systems>CommsGBU>Global Signaling Solutions>SN-SND: Tekelec EMS
 

In this Document

Goal

Solution

References

Created from <SR 3-12077167491>

Applies to:

Goal

We are currently working on New e5ms installation, where it was observer MYSQL & JAVA PROCESS IN E5MS SYSTEM IS USING MORE THAT 100 % OF CPU UTILIZATION,Due to which we suspect that the report generated using i-net clear report is generating error of "TIMEOUT-2003".
 

Solution

Your issue is related to Bug Bug 19307798 - innodb_buffer_pool of insufficient size for Measurement usage
Found in release 45.0 and fixed in release 46.0.0-460.7.0

The reason is that buffer pool is exhausted in E5MS due to small buffer size (only 8MB).
The error seen is below (this is from the customer system for today):
This issue becomes more prominent when E5MS is processing measurement data. In release 46 and above; the size of buffer has been increased to 4G.

[01 Feb 2016 15:51:54:633] FATAL: Exception while clearing Events for CLEAN_EVENT_INTERVAL at:
[01 Feb 2016 15:51:54:633] FATAL:
com.adventnet.nms.store.NmsStorageException: The total number of locks exceeds the lock table size

Query being executed when exception was thrown:

delete from Event where (TTIME < 1451643594226 )
  at com.adventnet.nms.store.relational.CommonEventToRelational.bulkDeleteOfEvents(CommonEventToRelational.java:510)
  at com.adventnet.nms.eventdb.CommonEventAPI.bulkDeleteOfEvents(CommonEventAPI.java:237)
  at com.adventnet.nms.eventdb.EventLog.cleanLogs(EventLog.java:483)
  at com.adventnet.nms.eventdb.EventMgr$CleanLogs.run(EventMgr.java:3903)
[
 

References

This solution has no attachment