5.8.5

Table Of Contents
Solution
1 Increase the number of FSDB savings threads by modifying the FSDBSaveThreads property in the
vcenter-ops\user\conf\analytics\advanced.properties file.
The default value for the FSDBSaveThreads property is 3. You can increase the value up to the number of
CPU cores on the host.
2 Use the FSDBHomeChanger tool to create additional mount points for the FSDB so that files are
distributed on multiple file systems.
3 Use the FSDBHomeChanger tool to move the FSDB home to a larger or faster drive.
4 Reduce the number of resources and metrics that are collected to reduce the demand on the drive.
Analytics Threshold Checking Overloaded
No more threshold checking threads are available.
Problem
The message Analytics Threshold Checking Overloaded appears in the Reason pane of the Alert Summary
page for the alert.
Cause
Threshold checking threads can become depleted when the vCenter Operations Manager Analytics server is
CPU bound or when the database access has reached its limit.
Solution
1 If the Analytics server has unused CPU cycles, increase the number of dynamic threshold processing
threads by changing the ThresholdProcessingThreads property in the vcenter-
ops\user\conf\analytics\advanced.properties file.
The default setting for the ThresholdProcessingThreads property is 10. The maximum value is 25, or the
number of CPU cores, whichever is higher.
2 Reduce the number of resources and metrics that are collected to reduce the demand on the
vCenter Operations Manager server host CPU.
3 If database access is causing the problem, check the latency between the vCenter Operations Manager
server and the database server.
Latency should be less than two milliseconds.
4 If necessary, upgrade the database server so that it can handle the load from
vCenter Operations Manager.
Collector Is Down
A heartbeat message was not received from a collector.
Problem
The message Collector is Down appears in the Reason pane of the Alert Summary page for the alert.
Cause
By default, each collector sends a heartbeat message to the controller or Web service every three seconds to
indicate that it is up and running. This alert indicates that a one of the collectors did not send a heartbeat for
five minutes.
VMware vCenter Operations Manager Administration Guide
120 VMware, Inc.