Cisco MCS-7825-H3-IPC1 Service Guide - Page 525

Chapter 6 Cisco Unified Serviceability Alarms and CiscoLog Messages Warning-Level Alarms Recommended Action Check the Cisco RIS Data Collector service parameter, Maximum Number of Processes and Threads, to see if the parameter has been set to a low value. If it has been, set the value higher or use the default value. Another possible action is that when a new Cisco product is integrated into Cisco Unified Communications Manager (Cisco Unified CM), new processes or threads are added to the system. Even in the normal process load situation, it's possible that the total number of processes and threads has exceeded the configured or default value of the Cisco RIS Data Collector service parameter, Maximum Number of Processes and Threads. Set that parameter to the maximum allowed value. You can also review the details of this alarm to check the ProcessWithMostThreads description and the ProcessWithMostInstances description to discover which processes have the most threads and the most instances. Determine whether these values are reasonable for this process; if not, contact the owner of the process for troubleshooting the reasons why the thread count or the number of process instances is so high. It is also possible that Cisco RIS Data Collector sent a false alarm, which would indicate a defect in the Cisco RIS Data Collector service. To determine if this is the cause of the alarm - after you have checked all the other errors described here - use RTMT to check the System object for performance counters Total Threads and Total Processes to confirm that the values in those counters do not exceed the value configured in the Cisco RIS Data Collector service parameter, Maximum Number of Processes and Threads. If the counters do not show a value that is higher than what is configured in the service parameter, restart Cisco RIS Data Collector service. If the alarm persists after restarting the service, go to Cisco Unified Serviceability and collect trace logs (Trace > Configuration) for Cisco Syslog, Cisco RIS Data Collector, Cisco AMC Service, and Cisco RIS Perfmon Logs and contact Cisco Technical Assistance Center (TAC) for detailed assistance. ThreadKillingError An error occurred when CMI tried to stop the CMI service. As a normal part of the process of stopping the CMI service, open threads are closed (killed). This alarm indicates that a timeout has occurred which means that the shutdown process is taking longer than expected, causing the operating system to return an error. History Cisco Unified Communications Release 8.0(1) Action Name changed from kThreadKillingError. Enum Definitions for MediaResourceType is updated. Cisco Unified Serviceability Alarm Definition Catalog CMIAlarmCatalog/CMI Severity WARNING Routing List Event Log SDI OL-22523-01 Cisco Unified Communications Manager Managed Services Guide 6-291