Cisco Cisco Unified Contact Center Enterprise 9.0(1)
Serviceability Best Practices Guide for Unified ICM/Unified CCE & Unified CCH
©2012 Cisco Systems, Inc.
104
8 Performance Counters
8.1
Platform Health Monitoring Counters
The following table lists the performance counters that you should watch on a regular basis to
determine the health of the contact center application.
determine the health of the contact center application.
Table 8-1: Performance Counters - Health Monitoring
Performance Object
Counter Name
(Instance)
(Instance)
Type
Units (Range)
Threshol
d
d
Green
Threshol
d
d
Yellow
Threshol
d
d
Red
Processor
% Processor Time
(_Total)
(_Total)
Int32
Percentage (0 -
100%)
100%)
< 50%
50% -
60%
60%
> 60%
(sustained
)
(sustained
)
Primary indicator of processor activity; displays the average percentage of CPU busy time observed during the
sample interval.
sample interval.
System
Processor Queue
Length
Length
Int32
# threads
< 2 *
#CPUs
#CPUs
-
>= 2 *
#CPUs
(sustained
)
#CPUs
(sustained
)
Number of threads in the processor queue waiting to be serviced. Microsoft states that Processor Queue Length
is OK up to 10 per CPU. This may be the case for non-real time applications but Unified CC performance is
impacted if this queue length is excessive for a sustained period of time. Timeouts are likely if the server
becomes CPU bound or a single application (or process) monopolizes the CPU.
is OK up to 10 per CPU. This may be the case for non-real time applications but Unified CC performance is
impacted if this queue length is excessive for a sustained period of time. Timeouts are likely if the server
becomes CPU bound or a single application (or process) monopolizes the CPU.
Memory
Available Bytes
Int32
Percentage (0 -
100%)
100%)
> 30%
20% -
30%
30%
< 20%
Amount of physical memory available to running processes; threshold values are a percentage of physical
memory. This is a snap shot–not a running average. Sustained samples below 20% (available) may be indicative
of a memory leak.
memory. This is a snap shot–not a running average. Sustained samples below 20% (available) may be indicative
of a memory leak.
Memory
Pages / sec
Int32
# page faults
< 10
>= 10
> 10
(sustained
)
(sustained
)
Pages/sec is the rate at which pages are read from or written to disk to resolve hard page faults. Excessive page
faults adversely impacts performance – root cause must be investigated.
faults adversely impacts performance – root cause must be investigated.
Physical Disk
Avg. Disk Queue
Length (_Total)
Length (_Total)
Float
Average #
read/write
requests
read/write
requests
< 1.5
-
>= 1.5
(sustained
)
(sustained
)
Average number of both read and write requests that were queued for the selected disk during the sample
interval.
interval.
Physical Disk
% Disk Time
(_Total)
(_Total)
Int32
Percentage (0 -
100%)
100%)
< 60%
60% -
80%
80%
> 80%
Percentage of elapsed time that the disk drive was busy servicing read or write requests.
Network Interface
Bytes Total/sec
Int32
Percentage (0 -
100%)
100%)
< 25%
25% -
30%
30%
> 30%
Rate at which bytes are sent and received over each network adapter. Threshold values are a percentage of
available bandwidth.
available bandwidth.
Network Interface
Output Queue
Length
Length
Int32
# packets in
queue
queue
0
1
> 1
(sustained
(sustained