• No se han encontrado resultados

Primeros espacios públicos en Chihuahua: alamedas, paseos y cementerios

Capítulo 1. Estado del arte “One city recycles the „left over‟ of the other into a sort of „second hand‟ urbanism.

1. Definiciones fundamentales

1.3 Espacio público: aproximaciones de estudio.

1.3.2 Primeros espacios públicos en Chihuahua: alamedas, paseos y cementerios

Teamcenter server monitoring metrics

You can configure the following metrics to provide specified levels of monitoring for specified threshold levels. Optionally, you can receive e-mail notification when specified metrics cross specified thresholds.

Metric Description

POMRetries Number of POM retries

Deadlocks Number of deadlock

VeryLongRunningQueries Number of very long running queries

DBConnectionLosses Number of database connection losses

Metric Description

SqlTripCount Number of SQL trips

DetailedSqlStats Number of SQL statistics

SqlTotalTime Total SQL entries

OmAllocations Number of object manager (OM) allocations since

the last call

OmCurrentAllocated Number of total allocations for the current OM state

OsMemoryTotal Total memory space used by the operating system

OsMemoryPeak" Peak memory space used by the operating system

OsBsmUndoPool Size of the BSM undo pool from the operating

system

BsmUndoPoolInUse Size of the BSM undo pool in use

PomLocks Number of POM locks

OmModelData Total of Object Manager (OM) model data

OmRollbackData Total of Object Manager (OM) rollback data

Grave Events Fatal or unexpected errors occurring in the server

Configure Teamcenter server monitoring using either: • The serverMonitorConfig.xml file.

For more information about using the XML file, seeConfigure monitoring with the serverMonitorConfig.xml file.

• The J2EE server manager administrative interface.

For more information about using the J2EE server manager administrative interface, seeAdministering monitoring options for Teamcenter servers.

Tip You should review all monitoring settings, ensuring the thresholds are set

correctly for your site.

If you do not know the optimum monitoring setting for any given critical event, set the value to Collect. Collect the data and review to determine normal activity levels. Then set notification values slightly higher than normal activity levels.

Tip The contents of the e-mail notifications are generated from the

TC_ROOT/pool_manager/serverMonitorConfigInfo.xml file. (This is a companion file to the serverMonitorConfig.xml file.) For a complete list of possible causes and recommended actions for Teamcenter server monitoring, see this file.

Configure monitoring with the serverMonitorConfig.xml file

1. Open the TC_ROOT/pool_manager/serverMonitorConfig.xml file. 2. Set the mode to one of the following:

Enables monitoring of all the metrics listed in the file.

Disable_Alerts

Enables monitoring of all the metrics listed in the file, but disables all notifications of critical events, regardless of individual notification settings on any metric.

Off

Disables monitoring of all the metrics listed in the file.

3. (Optional) To be notified when criteria reaches the specified threshold, specify from whom, to whom, and how frequently e-mail notification of critical events are sent by setting the following EmailResponder values.

You can specify more than one EmailResponder id.

All EmailResponder id values in all subsequent monitoring metrics in this file must match one of the EmailResponder id values set here.

EmailResponder id

Specify an identification for this e-mail responder. Multiple e-mail

responders can be configured, in which case, the identifiers must be unique.

protocol

Specify the e-mail protocol by which notifications are sent. The only supported protocol is SMTP.

hostAddress

Specify the server host from which the e-mail notifications are sent. In a large deployment (with multiple server managers, or the Web tier running on different hosts) the host address identifies the location of the critical events.

fromAddress

Specify the address from which the notification E-mails are sent.

toAddress

Specify the address to which the notification E-mails are sent.

suppressionPeriod

Specify the amount of time (in seconds) to suppress e-mail notification of critical events.

For more information, see the suppression period example inIntroduction to monitoring.

emailFormat

Specify the format in which the e-mail is delivered. Valid values are html and text.

4. (Optional) To be notified when criteria reaches the specified threshold, specify to whom, and to which file, critical events are logged by setting the following

All LoggerResponder values in all subsequent monitoring metrics in this file must match the LoggerResponder id value set here.

LoggerResponder id

Specify an identification for this logger responder. Multiple logger responders can be configured, in which case the identifiers must be unique.

logFileName

Specify the name of the file to which critical events are logged.

suppressionPeriod

Specify the amount of time (in seconds) to suppress logging of critical events to the log file.

For more information, see the suppression period example inIntroduction to monitoring.

5. Configure the criteria for a critical event for any of the metrics in the file by: a. Specifying a particular EmailResponder, if desired.

b. Specifying a particular LoggerResponder, if desired. c. Setting the metric’s monitoring mode to one of the following:

Collect

Collect metric data and display results in the MBean view (within the server manager administrative interface) for this metric.

This is the default setting.

Alert

Collect metric data, display results in the MBean view for this metric, and send e-mail notifications when critical events occur.

Off

No metric data is collected.

d. Setting the remaining values to specify criteria that must be met to initiate a critical event for the metric.

6. Save the file.

Server monitoring is enabled for the metrics you configured.

Sample serverMonitorConfig.xml code

In the following example, two EmailResponder elements are configured. E-mail notification of deadlock critical events are sent to [email protected], and e-mail notification of very long running queries are sent to [email protected].

<?xml version="1.0" encoding="UTF-8"?> <!--

Copyright 2011 Siemens Product Lifecycle Management Software Inc. All Rights Reserved. ================================================================================ Copyright 2011.

Siemens Product Lifecycle Management Software Inc. All Rights Reserved.

================================================================================ Copyright 2011 Siemens Product Lifecycle Management Software Inc. All Rights Reserved. -->

<!-- Server Health Monitoring Configuration -->

<ApplicationConfig id="TcServer" mode="Normal" version="1.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:noNamespaceSchemaLocation="healthMonitorV1.0.xsd"> <RespondersConfig> <EmailResponder id="EmailResponder1"> <protocol value="smtp"/> <hostAddress value="svc1smtp.company.com"/> <fromAddress value="[email protected]" /> <toAddress value="[email protected]" /> <suppressionPeriod value="4200"/> <emailFormat value="html"/> </EmailResponder> <EmailResponder id="EmailResponder2"> <protocol value="smtp"/> <hostAddress value="svc1smtp.company.com"/> <fromAddress value="[email protected]" /> <toAddress value="[email protected]" /> <suppressionPeriod value="4200"/> <emailFormat value="html"/> </EmailResponder> <LoggerResponder id="LoggerResponder1"> <logFileName value="ServerManagerMonitoring.log" /> <suppressionPeriod value="0"/> </LoggerResponder> </RespondersConfig> <MetricsConfig>

<Metric id="Deadlocks" maxEntries="50" metricType="integer" mode="Alert" name="Deadlocks">

<Threshold> <ThresholdValue

description="If number of deadlocks exceeds this limit notify" name="NumberofDeadLocks" value="10"/> </Threshold> <Responders> <ResponderRef id="EmailResponder1"/> </Responders> </Metric>

<Metric id="VeryLongRunningQueries" maxEntries="150"

metricType="integer" mode="Alert" name="VeryLongRunningQueries"> <ThresholdWithPeriod>

<ThresholdValue

description="If number of very long queries exceeds this limit notify" name="NoOfVeryLongRunningQueries" value="50"/>

<ThresholdPeriod

description="Periods for which very long queries will be monitored." name="TimePeriodinSec" value="600"/> </ThresholdWithPeriod> <Responders> <ResponderRef id="EmailResponder2"/> </Responders> </Metric> </MetricsConfig> </ApplicationConfig>

Configure monitoring with the server manager administrative interface

This procedures assumes you have the server manager administrative interface running.

For information about starting the interface, seeStart the J2EE administrative interface.

1. Under the Administer Pool-name servers monitoring heading, click

id=Server_Monitoring_Configurations.

This view lists the monitoring mode and all the metrics available for monitoring. 2. Set the Health_monitoring_mode value to one of the following:

Normal

Enables monitoring of all the metrics listed in the file.

Enables monitoring of all the metrics listed in the file, but disables all notifications of critical events, regardless of individual notification settings on any metric.

Off

Disables monitoring of all the metrics listed in the file. 3. In the same view, click the Administer pool-nameservers

monitoring:type=Configuration,id=EmailResponder1 value.

The EmailResponder view appears.

4. (Optional) To be notified when criteria reaches the specified threshold, specify from whom, to whom, and how frequently e-mail notification of critical events are sent by setting the following EmailResponder1 values.

All EmailResponder1 values in all child monitoring metrics must match the values set here.

fromAddress

Specify the address from which the notification E-mails are sent.

hostAddress

Specify the server host from which the e-mail notifications are sent. In a large deployment (with multiple server managers, or the Web tier running on different hosts) the host address identifies the location of the critical events.

suppressionPeriod

Specify the amount of time (in seconds) to suppress e-mail notification of critical events.

For more information, see the suppression period example inIntroduction to monitoring.

toAddress

Specify the address to which the notification E-mails are sent. You can specify multiple e-mail addresses, separated by commas.

5. Click Apply.

6. Click Back to Agent View.

7. Under the Administer pool-name servers monitoring heading, click

id=LoggerResponder1.

8. (Optional) To be notified when criteria reaches the specified threshold, specify to whom, and to which file, critical events are logged by setting the following

LoggerResponder values.

All LoggerResponder values in all child monitoring metrics must match the

LoggerResponder values set here.

Log_filename

Suppression_period

Specify the amount of time (in seconds) to suppress logging of critical events to the log file.

For more information, see the suppression period example inIntroduction to monitoring.

9. Click Apply.

10. Click Back to Agent View.

11. Configure monitoring of any of the Teamcenter server metrics listed under the

Administer Pool-name servers monitoring heading.

a. Click the desired metric.

For example: type=Configuration,id=Deadlocks.

b. Set the value for the Configure_mode attribute to one of the following:

Collect

Collect metric data and display results in the MBean view for this metric. This is the default setting.

Alert

Collect metric data, display results in the MBean view for this metric, and send e-mail notifications when critical events occur.

Off

No metric data is collected.

c. (Optional) To be notified when the criteria reaches the specified

threshold, specify the EmailResponder, LoggerResponder values for the

Configure_responder_Ids attribute.

By default, these values are set to EmailResponder1 and LoggerResponder1. If you have configured multiple EmailResponder IDs, make sure you specify the desired EmailResponder.

d. Set the remaining values to specify criteria that must be met to initiate a critical event for the metric.

e. After specifying values for each monitoring metric, click Apply.

Monitoring system alerts

Client applications can register their implementation of standard JMX notification listeners with the JMX monitoring system. When a monitoring system alert occurs, registered listeners a notification that contains the following information:

• Source

The object name that generated the notification. The client application uses this source to communicate with the component that raised the alert to request additional information about the alert.

• Sequence number

An incremental integer used to order notifications.

• A string that indicates the monitoring component that raised the alert in a Java component format, for example”

com.teamcenter.mld.healthmonitoring.ServerManager.threshold

• Message

A summary of the alert from the originating metric, for example:

Teamcenter Alert: Business Logic Server Crashes exceeded threshold.

• User data

A string containing the MetricID value.

The client can requested the following additional information about the alert:

AlertSummary data. This is information specific to the alert raised.

AlertSubject data. This is information specific to the alert raised.

AlertEventData data. This is information specific to the alert raised.

MetricID value. Defined in the webtierMonitorConfig.xml file.

MetricName value. Defined in the webtierMonitorConfig.xml file.

MetricDescription data. Defined in the webtierMonitorConfig.xml file.

PossibleCauses data. Defined in the webtierMonitorConfigInfo.xml file.

RecommendedActions data. Defined in the webtierMonitorConfigInfo.xml

file.

Automatic metric collection

You can enable automatic collection of metrics based on the occurrence of an alert. When the alert occurs, all events for the metric are captured and stored in memory. The maximum number of records to keep in memory is configured by the

maxEntries attribute in the webtierMonitorConfigInfo.xml file. After collection

is initiated, it remains in effect until you manually change the monitor mode for the metric to any other supported mode.

The following is a sample configuration for automatic collection of

DBConnectionLosses and Deadlock metrics when the POMRetries alert occurs: <MetricModeController id=" MetricModeController1">

<targetedMetrics>

<MetricId value="DBConnectionLosses"/> <MetricId value="DeadLock"/>

</targetedMetrics>

</MetricModeController>

<Metric name=" POMRetries” id=" POMRetries” > <Responders>

<ResponderRef id="MetricModeController1" /> </Responders>

</Metric >

If the mode for a metric is already set to Collect or Alert, subsequent alerts are ignored.

Automatic log level change

You can configure the a logger to automatically change log level to a specific value when an alert occurs. If multiple instances of a responder have different target levels for a logger, the logger is set the highest value (larger number) using the following order: 1. FATAL 2. ERROR 3. WARN 4. INFO 5. DEBUG 6. TRACE

If you specify a log level for a logger that has been adjusted due to an alert on a metric, your value supersedes the responder setting and clears any log level changes in queue due to the alert. The following is a sample configuration for automatic log level change to Debug for the LogLeverController1 responder with a duration of 1000 seconds: <LogLevelController id="LogLevelController1"> <targetedLevel value="Debug"/> <duration value ="1000"/> <targetedLoggers> <loggerName value="Teamcenter.pom"/> <loggerName value="Teamcenter.bom"/> </targetedLoggers> </LogLevelController>