|
4
|
Performance Counters
Collection
This section is for selecting the performance counters that are
exposed at the application level. Performance counters are used to
monitor the immediate health state of an application by collecting
a snapshot of configured performance counters at regular intervals.
By default, the 'Avg. Request Time' , 'Monitored Requests/sec' , '%
Exception Events/sec' and '% Performance
Events/sec' performance counters are all exposed.
- Avg. Request Time: The average
time (in milliseconds) it takes to load a web page of the monitored
web application. [default = 15000 ms]
- Monitored Requests/sec: The
frequency at which the requests are being made for the web-pages of
the monitored web-application. [default = 15 requests/sec]
- % Exception Events/sec: This
counter indicates the rate of total number of exceptions raised by
web-pages within the monitored application. [default = 15%]
- % Performance Events/sec: This
counter indicates the number of times a web page in the monitored
application exceeded the acceptable page load time per second.
[default = 20%]
- Interval - number Seconds/Minutes/Hours/Days. State
performance counters are used to monitor the immediate health state
of an application by collecting a snapshot of configured
performance counters at number
minute intervals [default = 5 minutes].
- Custom Counter List...: This
option allows you to add your own performance counters in addition
to those already exposed by default by the management pack. So, if
your application has your own custom performance counters, or you
would like to use the MP to continue monitoring a particular set of
performance counters that you typically use, add them via the
Custom Counter List... option.
To Add or Edit Custom Performance
Counters: Select either Add
or Edit to open, add to or
modify the list of custom performance counters.
|
Specify the performance counter
|
|
a
|
Either enter the display name, object, counter, and instance
into the text field, or use Browse... to select the appropriate counter,
which will automatically fill in the text fields.
- Display name: The rule name for
collecting the custom performance counter (without aggregation) for
displaying in the Operations Console
- Object: The Object (category)
name for the performance counter.
- Counter: The name of the
performance aspect to be measured within the object category.
- Instance: The name of the
unique instance of the performance counter.
- Include all instances for the selected
counter: Although some objects (such as Memory and Server)
have only a single performance object instance, some performance
objects can have multiple instances. If an object has multiple
instances, you can add counters to track statistics for all
instances at once. Enable this checkbox to aggregate values for
this counter from across all discovered computers into a single
reported value. Disable this checkbox to only collect the selected
instance.
 If
the counters category does not support instances such as Memory or
Server, leave the instance text box empty.
To Browse and Select a Performance
Counter: Click Browse... to
look through all system counters.
|
Browse or enter computer name
- Computer: Enter the name of the
computer, or click ... to make a selection
- Performance Counter
- Object: Select the object
(category) from the dropdown list
- Select counter from list:
Select the counter
- Select instance from list:
Select the instance of the counter, or check the 'All Instances'
checkbox below
- Explain...: Display the
description for the counter in a popup window.
- All Instances: This checkbox is
always unchecked and disabled. Use Include
all instances for the selected counter from the parent
screen.
|
|
|
Aggregation settings
|
|
b
|
This section allows users to configure how data that is
aggregated from across multiple computer sources will be displayed
in the KPI dashboard.
- Aggregation Counter name: The
default name of the aggregation counter as it will appear on the
KPI dashboard. The name created by combining the name of the
selected counter and selected aggregation function. This name
may be changed by the user.
- Aggregation Function: This
dropdown allows users to select the function to use when
calculating the aggregation counter. Functions include calculating
the aggregated value based on the Sum, Average, Minimum or Maximum
values.
- Show trend in the dashboard:
When this checkbox is enabled, trends for the aggregated
performance counter will be shown in the KPI dashboard.
|
To Delete Custom Performance
Counters: Select a counter from the list and click
Delete
To Change the Order that Custom
Performance Counters will appear in the Trend Reports:
Select a counter from the list and click Move Up or Move
Down
Collect Alerts by Event
Type
The management pack reports problems in two ways: via problem
based alerts and via the application state. For each unique
problem signature, the MP generates a unique alert inside of
Operations Manager. This section allows user to decide
whether or not to generate problem based alerts, and which type of
problems based alerts should be generated.
- Exception: Select this checkbox
to enable rules to collect Exception alerts, or deselect to
disable the collection rules. [default=selected]
- Performance: Select this
checkbox to enable rules to collect Performance alerts, or
deselect to disable the collection rules. [default=selected]
Monitors
The Management Pack monitors at both the overall application
health level and at the business transaction level. The overall
application health state can be monitored with respect to the
number of exception and performance events and application
load.
- The % Exception Events/sec
exceeds number [default =
15 percent]
- The Average Request Time
exceeds number [default =
15000 milliseconds]
To Add or Edit Application Load
Monitors: Select either Add
or Edit.
Several
monitors of the same type can be added, but the thresholds for
these monitors must be different.
|
Select the monitor type from the list
(available for 'Add' only)
|
|
a
|
Select from the following list of available monitor types which
can be added for applications or transactions:
The Average Request Time exceeds
threshold: The instance state becomes unhealthy if the
average request execution time per predefined escalation time
period exceeds the escalation threshold. Application performance
alerts are generated when the execution time exceeds the configured
Alerting Threshold. Excessively long Network request times may be
due to unbalanced Network. Long Server processing may be due to
complex or inefficient server code. Page Loading Time performance
issues may occur because of either large page size or a large
number of JavaScript s. Long Window.OnLoad performance issues may
result from long JavaScript handlers execution. JavaScript
performance issues may result from long JavaScript handlers
execution.
The Number of Monitored Requests/sec
exceeds threshold: The instance state becomes unhealthy if
the total number of monitored requests per predefined escalation
time period exceeds an escalation threshold (i.e. the frequency at
which the requests are being made for the web-pages of the
monitored web-application). This counter is only available when
AVIcode Operations Monitoring and Client Side monitoring are
enabled. Therefore this counter shows the number of requests that
have a duration greater than the Application Alerting Threshold.
Excessively long Network request times may be due to unbalanced
Network. Long Server processing may be due to complex or
inefficient server code. Page Loading Time performance issues may
occur because of either large page size or a large number of
JavaScript s. Long Window.OnLoad performance issues may result from
long JavaScript handlers execution. JavaScript performance issues
may result from long JavaScript handlers execution.
The % Performance Events/sec exceeds
threshold: The instance state becomes unhealthy if the
total number of performance events per predefined escalation time
period exceeds an escalation threshold. Application performance
alerts are generated when the number of performance events exceeds
the configured monitor threshold. Excessively long Network request
times may be due to unbalanced Network. Long Server processing may
be due to complex or inefficient server code. Page Loading Time
performance issues may occur because of either large page size or a
large number of JavaScript s. Long Window.OnLoad performance issues
may result from long JavaScript handlers execution. JavaScript
performance issues may result from long JavaScript handlers
execution.
The % Exception Events/sec exceeds
threshold: The instance state becomes unhealthy if the
total number of exception events per predefined escalation time
period exceeds an escalation threshold. Application availability
alerts are generated when the number of exception events exceeds
the configured monitor threshold. Therefore this counter indicates
the rate of total number of exceptions raised by web-pages within
the monitored application. Application failure alerts arise from
JavaScript exceptions. To resolve these exceptions, check the
exception type and the function that the exception was thrown from.
Resolution of these types of problems typically requires access to
application source code.
|
|
Monitor properties (for both Add and Edit)
|
|
b
|
Monitor Settings: The read-only
name of the selected monitor
- Specify threshold value to
match:
Threshold:
Number percent. This number indicates how many performance or
exception events (as a percentage of all monitored requests) can
occur per interval before affecting the health state. Available
only for '% Performance Event' [default = 20 percent] and '%
Exception Event' [default = 15 percent] monitors.
Threshold:
Number milliseconds. This number indicates the most number of
milliseconds the average request time can last per interval before
affecting the health state. Available only for 'The Average
Request Time' monitor [default = 2000 ms].
Threshold:
Number requests/sec. This number indicates the maximum monitored
requests that can be made per interval before affecting the health
state. Available only for 'The Number of Monitored
Requests/sec' monitor [default = 15 requests/sec].
- Specify Health State:
Health State: Select whether the
monitor will go to a Warning or Error health state when the
specified threshold is met for the specified interval [default =
Error].
- Specify the interval: Interval
number Seconds/Minutes/Hours/Days.
This number indicates the
frequency with which the Warning and Error thresholds will be
checked. For example, if the Warning threshold is 5, then a
state change will take place if 5 events occur within the
interval. It also indicates the frequency of the Monitor Rule check
[default = 5 minutes].
- Alert Properties: The alert
properties are displayed when an alert is generated in response to
the monitor changing state. The properties include: Alert Name (Provide a unique, descriptive name
for the alert to be generated), and Alert
Description (Provide a description of the possible causes
of, and resolutions to, this alert). The alert description can
contain any number of parameters formatted as %ParameterName%, that
will be replaced with actual values when alert is generated.
Possible parameters include: %Threshold%, %InstanceName%, %Value%.
The alert name can not contain parameters.
- Automatically resolve the alert when
the monitor returns to a healthy state: Select this checkbox
to have the MP automatically change alert resolution state to
'Closed' when the monitor returns to a healthy state [default
= selected].
|
To Delete Monitors: select a
monitor from the list and click Delete
|
|
5
|
The Management Pack not only monitors overall application
health, but also monitors at the business transaction level. So, in
addition to monitoring things like the overall volume of activity
on an applications, you may have, for example, 5 or 10 top
business transactions that you would like to monitor. In this
case you would define your own monitoring rules to see the volume
of traffic on those transactions or the average overall application
response time, and you want to have the ability to define monitors
and health rules for those 5 or 10 critical transactions. For each
transaction, we can expose the same performance counters as were
exposed for the overall application, like average response time or
number of events.
The discovery interval for transactions is defined in section 3
above.
Transactions: This list displays
the current list of monitored transactions. New Web Pages can be
added to the list .
To Add a Web Page: Click
Add web page.
|
ASP.NET Page Settings
|
|
a
|
- Transaction Name: A user
defined string to describe a transaction to a web page
- ASP.NET Page: The virtual path
to the selected ASPX page, for example "/Virtual
Dir/mypage.aspx"
To Browse and Select an ASP.NET
Page: Click Browse...
|
Specify Account for Remote
Access
- Use current user checkbox
[default = selected]
- Account Name [default =
disabled]
- Password [default =
disabled]
Select Computer:
- Enter the name of the computer you want to browse assemblies on
or choose from a computer list. Leave this field empty for using
local computer ('...')
|
|
Performance Counters
- Avg. Request Time: The average
time (in milliseconds) that the monitored page has been executed.
[default = 10000 ms]
- Monitored Requests/sec: The
total number of calls of the monitored page that are being executed
each second. [default = 15%]
- % Exception Events/sec: The
rate of exceptions raised in the monitored page to whole calls of
monitored page. [default = 5%]
- % Performance Events/sec: The
rate of page calls that have exceeded their threshold to whole
calls of monitored page. [default = 20%]
- Interval: Number Seconds/Minutes/Hours/Days. State
performance counters are used to monitor the immediate health state
of an application by collecting a snapshot of configured
performance counters at number
minute intervals [default = 5 minutes].
Collect Alerts by Event Type
The management pack reports problems in two ways: via problem
based alerts and via the application state. For each unique problem
signature, the MP generates a unique alert inside of Operations
Manager. This section allows user to decide whether or not to
generate problem based alerts, and which type of problems based
alerts should be generated.
- Exception: This field tells the
Management Pack whether or not you want Operations Manager to
report the health of your .NET application by the frequency of
exception problems. Select this checkbox to enable rules
to collect Exception alerts, or deselect to disable the
collection rules [default = selected].
- Performance: This field tells
the Management Pack whether or not you want Operations Manager to
report the health of your .NET application by the frequency of
performance problems. Select this checkbox to enable rules
to collect Performance alerts, or deselect to disable the
collection rules [default = selected].
Monitors
To Add or Edit Monitors: Select
either Add or Edit.
|
Select the monitor type from the list
Select from the following list of available monitor types:
The Average Request Time exceeds
threshold: The instance state becomes unhealthy if the
average request execution time per predefined escalation time
period exceeds the escalation threshold. Application performance
alerts are generated when the execution time exceeds the configured
Alerting Threshold. Excessively long Network request times may be
due to unbalanced Network. Long Server processing may be due to
complex or inefficient server code. Page Loading Time performance
issues may occur because of either large page size or a large
number of JavaScript s. Long Window.OnLoad performance issues may
result from long JavaScript handlers execution. JavaScript
performance issues may result from long JavaScript handlers
execution.
The Number of Monitored Requests/sec
exceeds threshold: The instance state becomes unhealthy if
the total number of monitored requests per predefined escalation
time period exceeds an escalation threshold (i.e. the frequency at
which the requests are being made for the web-pages of the
monitored web-application). This counter is only available when
AVIcode Operations Monitoring and Client Side monitoring are
enabled. Therefore this counter shows the number of requests that
have a duration greater than the Application Alerting Threshold.
Excessively long Network request times may be due to unbalanced
Network. Long Server processing may be due to complex or
inefficient server code. Page Loading Time performance issues may
occur because of either large page size or a large number of
JavaScript s. Long Window.OnLoad performance issues may result from
long JavaScript handlers execution. JavaScript performance issues
may result from long JavaScript handlers execution.
The % Performance Events/sec exceeds
threshold: The instance state becomes unhealthy if the
total number of performance events per predefined escalation time
period exceeds an escalation threshold. Application performance
alerts are generated when the number of performance events exceeds
the configured monitor threshold. Excessively long Network request
times may be due to unbalanced Network. Long Server processing may
be due to complex or inefficient server code. Page Loading Time
performance issues may occur because of either large page size or a
large number of JavaScript s. Long Window.OnLoad performance issues
may result from long JavaScript handlers execution. JavaScript
performance issues may result from long JavaScript handlers
execution.
The % Exception Events/sec exceeds
threshold: The instance state becomes unhealthy if the
total number of exception events per predefined escalation time
period exceeds an escalation threshold. Application availability
alerts are generated when the number of exception events exceeds
the configured monitor threshold. Therefore this counter indicates
the rate of total number of exceptions raised by web-pages within
the monitored application. Application failure alerts arise from
JavaScript exceptions. To resolve these exceptions, check the
exception type and the function that the exception was thrown from.
Resolution of these types of problems typically requires access to
application source code.
Monitor properties (for both Add and Edit)
- Monitor Settings: The read-only
name of the selected monitor
- Specify threshold value to
match:
Threshold:
Number percent. This number indicates how many performance or
exception events (as a percentage of all monitored requests) can
occur per interval before affecting the health state. Available
only for '% Performance Event' [default = 10 percent] and '%
Exception Event' monitors [default = 5 percent].
Threshold:
Number milliseconds. This number indicates the most number of
milliseconds the average request time can last per interval before
affecting the health state. Available only for 'The Average
Execution Time' monitor. [default = 10000 ms]
Threshold:
Number calls/sec. This number indicates the maximum monitored calls
that can be made per interval before affecting the health
state. Available only for 'The Number of Calls/sec'
monitor. [default = 15 calls/sec]
- Specify Health State:
Health State: Select whether the
monitor will go to a Warning or Error health state when the
specified threshold is met for the specified interval [default =
Error].
- Specify the interval:
Interval number Seconds/Minutes/Hours/Days. This number indicates the frequency with which the
Warning and Error thresholds will be checked. For example, if
the Warning threshold is 5, then a state change will take place
if 5 events occur within the interval. It also indicates the
frequency of the Monitor Rule check [default = 5 minutes].
- Alert Properties: The alert
properties are displayed when an alert is generated in response to
the monitor changing state. The properties include: Alert Name (Provide a unique, descriptive name
for the alert to be generated), and Alert
Description (Provide a description of the possible causes
of, and resolutions to, this alert). The alert description can
contain any number of parameters formatted as %ParameterName%, that
will be replaced with actual values when alert is generated.
Possible parameters include: %Threshold%, %InstanceName%, %Value%.
The alert name can not contain parameters.
- Automatically resolve the alert when
the monitor returns to a healthy state: Select this checkbox
to have the MP automatically change alert resolution state to
'Closed' when the monitor returns to a healthy state
[default = selected].
|
|
"Advanced Configuration"
"Authoring View"', right click
the 
"Management Pack Templates" folder
and select 'Add Monitoring Wizard.' This action will open an
installation wizard which will guide you through the activation
process.
"Enterprise ASP.NET Client Side
Application"