This chapter describes the tools you can use to monitor a broker and how you can get metrics data. The chapter has the following sections:
Reference information on specific metrics is available in Metrics Information Reference
The broker includes components for monitoring and diagnosing application and broker performance. These include the components and services shown in the following figure:
Broker code that logs broker events.
A metrics generator that provides.
The metrics generator provides information about broker activity, such as message flow in and out of the broker, the number of messages in broker memory and the memory they consume, the number of open connections, and the number of threads being used. The boolean broker property
imq.metrics.enabled controls whether such information is logged and the
imq.metrics.interval property specifies how often metrics information is generated.
A logger component that writes out information to a number of output channels.
A comprehensive set of Java Management Extensions (JMX) MBeans that expose broker resources using the JMX API
Support for the Java ES Monitoring Framework
A metrics message producer that sends JMS messages containing metrics information to topic destinations for consumption by JMS monitoring clients.
Broker properties for configuring the monitoring services are listed under Monitoring Properties.
There are five tools (or interfaces) for monitoring Message Queue information, as described briefly below:
Log files provide a long-term record of metrics data, but cannot easily be parsed.
The Command Utility (
imqcmd metrics) lets you interactively sample information tailored to your needs, but does not provide historical information or allow you to manipulate the data programmatically.
The Java Management Extensions (JMX) Administration API lets you perform broker resource configuration and monitoring operations programmatically from within a Java application. You can write your own JMX administration application or use the standard Java Monitoring and Management Console (
The Sun Java Enterprise System Monitoring Framework (JESMF) and Monitoring Console offers a common, Web-based graphical interface shared with other Java ES components, but can monitor only a subset of all Message Queue entities and operations.
The Message-based Monitoring API lets you extract metrics information from messages produced by the broker to metrics topic destinations. However, to use it, you must write a Message Queue client application to capture, analyze, and display the metrics data.
The following tabel compares the different tools.
|Metrics Monitoring Tool||Benefits||Limitations|
Command Utility (
JMX Administration API
Java ES Monitoring Console
Message-based Monitoring API
In addition to the differences shown in the table, each tool gathers a somewhat different subset of the metrics information generated by the broker. For information on which metrics data is gathered by each monitoring tool, see Metrics Information Reference.
The Message Queue Logger takes information generated by broker code, a debugger, and a metrics generator and writes that information to a number of output channels: to standard output (the console), to a log file, and, on Solaris platforms, to the
syslog daemon process. You can specify the type of information gathered by the Logger as well as the type of information the Logger writes to each of the output channels. For example, you can specify that you want metrics information written out to a log file.
This section describes the configuration and use of the Logger for monitoring broker activity. It includes the following topics:
imq.log.file.filename broker properties identify the log file to use and the
imq.log.console.stream property specifies whether console output is directed to
imq.log.level property controls the categories of metric information that the Logger gathers:
INFO. Each level includes those above it, so if you specify, for example,
WARNING as the logging level, error messages will be logged as well.
There is also an
imq.destination.logDeadMsgs property that specifies whether to log entries when dead messages are discarded or moved to the dead message queue.
imq.log.file.output properties control which of the specified categories the Logger writes to the console and the log file, respectively. In this case, however, the categories do not include those above them; so if you want, for instance, both errors and warnings written to the log file and informational messages to the console, you must explicitly set
On Solaris platforms another property,
imq.log.syslog.output, specifies the categories of metric information to be written to the
In the case of a log file, you can specify the point at which the file is closed and output is rolled over to a new file. Once the log file reaches a specified size (
imq.log.file.rolloverbytes) or age (
imq.log.file.rolloversecs), it is saved and a new log file created.
See Monitoring Properties for additional broker properties related to logging and subsequent sections for details about how to configure the Logger and how to use it to obtain performance information.
A logged message consists of a time stamp, a message code, and the message itself. The volume of information included varies with the logging level you have set. The broker supports three logging levels:
WARNING , and
INFO (see Table 13-2). Each level includes those above it (for example,
Serious problems that could cause system failure
Conditions that should be heeded but will not cause system failure
Metrics and other informational messages
The default logging level is
INFO, so messages at all three levels are logged by default. The following is an example of an
[13/Sep/2000:16:13:36 PDT] [B1004]: Starting the broker service using tcp [25374,100] with min threads 50 and max threads of 500
You can change the time zone used in the time stamp by setting the broker configuration property
imq.log.timezone (see Table 17-13).
A broker is automatically configured to save log output to a set of rolling log files. The log files are located in a directory identified by the instance name of the associated broker:
For a broker whose life cycle is controlled by GlassFish Server, the log files are located in a subdirectory of the domain directory for the domain for which the broker was started:
The log files are simple text files. The system maintains nine backup files named as follows, from earliest to latest:
log.txt log_1.txt log_2.txt … log_9.txt
By default, the log files are rolled over once a week. You can change this rollover interval, or the location or names of the log files, by setting appropriate configuration properties:
To change the directory in which the log files are kept, set the property
imq.log.file.dirpath to the desired path.
To change the root name of the log files from
log to something else, set the
To change the frequency with which the log files are rolled over, set the property
See Table 17-13 for further information on these properties.
Log-related properties are described in Table 17-13.
Set the logging level.
Set the output channel (file, console, or both) for one or more logging categories.
If you log output to a file, configure the rollover criteria for the file.
You complete these steps by setting Logger properties. You can do this in one of two ways:
Change or add Logger properties in the
config.properties file for a broker before you start the broker.
Specify Logger command line options in the
imqbrokerd command that starts the broker. You can also use the broker option
-D to change Logger properties (or any broker property).
Options passed on the command line override properties specified in the broker instance configuration files. The following
imqbrokerd options affect logging:
Logging interval for broker metrics, in seconds
Logging level (
Silent mode (no logging to console)
Log all messages to console
The following sections describe how you can change the default configuration in order to do the following:
Change the output channel (the destination of log messages)
Change rollover criteria
By default, error and warning messages are displayed on the terminal as well as being logged to a log file. (On Solaris, error messages are also written to the system's
You can change the output channel for log messages in the following ways:
To have all log categories (for a given level) output displayed on the screen, use the
-tty option to the
To prevent log output from being displayed on the screen, use the
-silent option to the
imq.log.file.output property to specify which categories of logging information should be written to the log file. For example,
imq.log.console.output property to specify which categories of logging information should be written to the console. For example,
On Solaris, use the
imq.log.syslog.output property to specify which categories of logging information should be written to Solaris
syslog. For example,
Before changing Logger output channels, you must make sure that logging is set at a level that supports the information you are mapping to the output channel. For example, if you set the logging level to
ERROR and then set the
imq.log.console.output property to
WARNING, no messages will be logged because you have not enabled the logging of
There are two criteria for rolling over log files: time and size. The default is to use a time criteria and roll over files every seven days.
To change the time interval, you need to change the property
imq.log.file.rolloversecs. For example, the following property definition changes the time interval to ten days:
To change the rollover criteria to depend on file size, you need to set the
imq.log.file.rolloverbytes property. For example, the following definition directs the broker to rollover files after they reach a limit of 500,000 bytes
If you set both the time-related and the size-related rollover properties, the first limit reached will trigger the rollover. As noted before, the broker maintains up to nine rollover files.
You can set or change the log file rollover properties when a broker is running. To set these properties, use the
imqcmd update bkr command.
This section describes the procedure for using broker log files to report metrics information. For general information on configuring the Logger, see Configuring and Using Broker Logging.
Configure the broker's metrics generation capability:
Generation of metrics for logging is turned on by default.
Set the metrics generation interval to a convenient number of seconds.
This value can be set in the
config.properties file or using the
-metrics interval command line option when starting up the broker.
Confirm that the Logger gathers metrics information:
This is the default value. This value can be set in the
config.properties file or using the
-loglevel level command line option when starting up the broker.
Confirm that the Logger is set to write metrics information to the log file:
This is the default value. It can be set in the
Start up the broker.
The following shows sample broker metrics output to the log file:
[21/Jul/2004:11:21:18 PDT] Connections: 0 JVM Heap: 8323072 bytes (7226576 free) Threads: 0 (14-1010) In: 0 msgs (0bytes) 0 pkts (0 bytes) Out: 0 msgs (0bytes) 0 pkts (0 bytes) Rate In: 0 msgs/sec (0 bytes/sec) 0 pkts/sec (0 bytes/sec) Rate Out: 0 msgs/sec (0 bytes/sec) 0 pkts/sec (0 bytes/sec)
For reference information about metrics data, see Metrics Information Reference.
You can monitor physical destinations by enabling dead message logging for a broker. You can log dead messages whether or not you are using a dead message queue.
If you enable dead message logging, the broker logs the following types of events:
A physical destination exceeded its maximum size.
The broker removed a message from a physical destination, for a reason such as the following:
The destination size limit has been reached.
The message time to live expired.
The message is too large.
An error occurred when the broker attempted to process the message.
If a dead message queue is in use, logging also includes the following types of events:
The broker moved a message to the dead message queue.
The broker removed a message from the dead message queue and discarded it.
The following is an example of the log format for dead messages:
[29/Mar/2006:15:35:39 PST] [B1147]: Message 8-188.8.131.52(e7:6b:dd:5d:98:aa)- 35251-1143675279400 from destination Q:q0 has been placed on the DMQ because [B0053]: Message on destination Q:q0 Expired: expiration time 1143675279402, arrival time 1143675279401, JMSTimestamp 1143675279400
Dead message logging is disabled by default. To enable it, set the broker attribute
A Message Queue broker can report metrics of the following types:
Java Virtual Machine (JVM) metrics. Information about the JVM heap size.
Brokerwide metrics. Information about messages stored in a broker, message flows into and out of a broker, and memory use. Messages are tracked in terms of numbers of messages and numbers of bytes.
Connection Service metrics. Information about connections and connection thread resources, and information about message flows for a particular connection service.
Destination metrics. Information about message flows into and out of a particular physical destination, information about a physical destination's consumers, and information about memory and disk space usage.
imqcmd command can obtain metrics information for the broker as a whole, for individual connection services, and for individual physical destinations. To obtain metrics data, you generally use the
metrics subcommand of
imqcmd. Metrics data is written at an interval you specify, or the number of times you specify, to the console screen.
You can also use the
query subcommand to view similar data that also includes configuration information. See imqcmd query for more information.
|Subcommand Syntax||Metrics Data Provided|
metrics bkr [-b hostName:portNumber] [-m metricType] [-int interval] [-msp numSamples]
Displays broker metrics for the default broker or a broker at the specified host and port.
metrics svc -n serviceName [-b hostName:portNumber] [-m metricType] [-int interval] [-msp numSamples]
Displays metrics for the specified service on the default broker or on a broker at the specified host and port.
metrics dst -t destType -n destName [-b hostName:portNumber] [-m metricType] [-int interval] [-msp numSamples]
Displays metrics information for the physical destination of the specified type and name.
Specifies the hostname and port of the broker for which metrics data is reported. The default is
Literal IP addresses as host names: You can use a literal IPv4 or IPv6 address as a host name. If you use a literal IPv6 address, its format must conform to RFC2732, Format for Literal IPv6 Addresses in URL's.
Specifies the interval (in seconds) at which to display the metrics. The default is 5 seconds.
Specifies the type of metric to display:
Specifies the number of samples displayed in the output. The default is an unlimited number (infinite).
Specifies the name of the physical destination (if any) for which metrics data is reported. There is no default.
Specifies the connection service (if any) for which metrics data is reported. There is no default.
Specifies the type (queue or topic) of the physical destination (if any) for which metrics data is reported. There is no default.
This section contains examples of output for the
imqcmd metrics subcommand. The examples show brokerwide, connection service, and physical destination metrics.
To get the rate of message and packet flow into and out of the broker at 10 second intervals, use the
metrics bkr subcommand:
imqcmd metrics bkr -m rts -int 10 -u admin
This command produces output similar to the following (see data descriptions in Table 21-2):
-------------------------------------------------------- Msgs/sec Msg Bytes/sec Pkts/sec Pkt Bytes/sec In Out In Out In Out In Out -------------------------------------------------------- 0 0 27 56 0 0 38 66 10 0 7365 56 10 10 7457 1132 0 0 27 56 0 0 38 73 0 10 27 7402 10 20 1400 8459 0 0 27 56 0 0 38 73
To get cumulative totals for messages and packets handled by the jms connection service, use the
metrics svc subcommand:
imqcmd metrics svc -n jms -m ttl -u admin
This command produces output similar to the following (see data descriptions in Table 21-3):
------------------------------------------------- Msgs Msg Bytes Pkts Pkt Bytes In Out In Out In Out In Out ------------------------------------------------- 164 100 120704 73600 282 383 135967 102127 657 100 483552 73600 775 876 498815 149948
To get metrics information about a physical destination, use the
metrics dst subcommand:
imqcmd metrics dst -t q -n XQueue -m ttl -u admin
This command produces output similar to the following (see data descriptions in Table 21-4):
----------------------------------------------------------------------------- Msgs Msg Bytes Msg Count Total Msg Bytes (k) Largest In Out In Out Current Peak Avg Current Peak Avg Msg (k) ----------------------------------------------------------------------------- 200 200 147200 147200 0 200 0 0 143 71 0 300 200 220800 147200 100 200 10 71 143 64 0 300 300 220800 220800 0 200 0 0 143 59 0
To get information about a physical destination's consumers, use the following
metrics dst subcommand:
imqcmd metrics dst -t q -n SimpleQueue -m con -u admin
This command produces output similar to the following (see data descriptions in Table 21-4):
------------------------------------------------------------------ Active Consumers Backup Consumers Msg Count Current Peak Avg Current Peak Avg Current Peak Avg ------------------------------------------------------------------ 1 1 0 0 0 0 944 1000 525
The syntax and options of
imqcmd query are shown in Table 13-5 along with a description of the metrics data provided by the command.
|Subcommand Syntax||Metrics Data Provided|
query bkr [-b hostName:portNumber]
Information on the current number of messages and message bytes stored in broker memory and persistent store (see Viewing Broker Information).
query svc -n serviceName [-b hostName:portNumber]
Information on the current number of allocated threads and number of connections for a specified connection service (see Viewing Connection Service Information).
query dst -t destType -n destName [-b hostName:portNumber]
Information on the current number of producers, active and backup consumers, and messages and message bytes stored in memory and persistent store for a specified destination (see Viewing Physical Destination Information).
Because of the limited metrics data provided by
imqcmd query , this tool is not represented in the tables presented in Metrics Information Reference.
The broker implements a comprehensive set of Java Management Extensions (JMX) MBeans that represent the broker's manageable resources. Using the JMX API, you can access these MBeans to perform broker configuration and monitoring operations programmatically from within a Java application.
In this way, the MBeans provide a Java application access to data values representing static or dynamic properties of a broker, connection, destination, or other resource. The application can also receive notifications of state changes or other significant events affecting the resource.
JMX-based administration provides dynamic, fine grained, programmatic access to the broker. You can use this kind of administration in a number of ways.
You can include JMX code in your JMS client application to monitor application performance and, based on the results, to reconfigure the Message Queue resources you use to improve performance.
You can write JMX client applications that monitor the broker to identify use patterns and performance problems, and you can use the JMX API to reconfigure the broker to optimize performance.
You can write a JMX client application to automate regular maintenance tasks.
You can write a JMX client application that constitutes your own version of the Command utility (
imqcmd), and you can use it instead of
You can use the standard Java Monitoring and Management Console (
jconsole) that can provide access to the broker's MBeans.
For information on JMX infrastructure and configuring the broker's JMX support, see JMX Support. To manage a Message Queue broker using the JMX architecture, see the Oracle GlassFish Server Message Queue Developer's Guide for JMX Clients.
Message Queue supports the Sun Java System Monitoring Framework (JESMF), which allows Java Enterprise System (Java ES) components to be monitored using a common graphical interface, the Sun Java System Monitoring Console. Administrators can use the Monitoring Console to view performance statistics, create rules for automatic monitoring, and acknowledge alarms. If you are running Message Queue along with other Java ES components, you may find it more convenient to use a single interface to manage all of them.
The Java ES Monitoring Framework defines a common data model, the Common Monitoring Model (CMM), to be used by all Java ES component products. This model enables a centralized and uniform view of all Java ES components. Message Queue exposes the following objects through the Common Monitoring Model:
The installed product
The broker instance name
The broker Port Mapper
Each connection service
Each physical destination
The persistent data store
The user repository
Each of these objects is mapped to a CMM object whose attributes can be monitored using the Java ES Monitoring Console. The reference tables in JES Monitoring Framework Reference identify those attributes that are available for JESMF monitoring. For detailed information about the mapping of Message Queue objects to CMM objects, see the Sun Java Enterprise System Monitoring Guide.
To enable JESMF monitoring, you must do the following:
Enable and configure the Monitoring Framework for all of your monitored components, as described in the Sun Java Enterprise System Monitoring Guide.
Install the Monitoring Console on a separate host, start the master agent, and then start the Web server, as described in the Sun Java Enterprise System Monitoring Guide.
Using the Java ES Monitoring Framework will not affect broker performance, because all the work of gathering metrics is done by the Monitoring Framework, which pulls data from the broker's existing data monitoring infrastructure.
For information on metric information provided by the Java ES Monitoring Framework, see JES Monitoring Framework Reference.
Message Queue provides a Metrics Message Producer, which receives information from the Metrics Generator at regular intervals and writes the information into metrics messages,. The Metrics Message Producer then sends these messages to one of a number of metric topic destinations, depending on the type of metric information contained in the messages.
You can access this metrics information by writing a client application that subscribes to the metrics topic destinations, consumes the messages in these destinations, and processes the metrics information contained in the messages. This allows you to create custom monitoring tools to support messaging applications. For details of the metric quantities reported in each type of metrics message, see "Using the Metrics Monitoring API" in Oracle GlassFish Server Message Queue Developer's Guide for Java Clients.
There are five metrics topic destinations, whose names are shown in Table 13-6, along with the type of metrics messages delivered to each destination.
Java Virtual Machine metrics
List of destinations and their types
Destination metrics for queue queueName
Destination metrics for topic topicName
The broker properties
imq.metrics.topic.interval control, respectively, whether messages are sent to metric topic destinations and how often. The
imq.metrics.topic.persist properties specify the lifetime of such messages and whether they are persistent.
Besides the information contained in the body of a metrics message, the header of each message includes properties that provide the following additional information:
The message type
The address (host name and port number) of the broker that sent the message
The time the metric sample was taken
These properties are useful to client applications that process metrics messages of different types or from different brokers.
This section describes the procedure for using the message-based monitoring capability to gather metrics information. The procedure includes both client development and administration tasks.
Write a metrics monitoring client.
See the Oracle GlassFish Server Message Queue Developer's Guide for Java Clients for instructions on programming clients that subscribe to metrics topic destinations, consume metrics messages, and extract the metrics data from these messages.
Configure the broker's Metrics Message Producer by setting broker property values in the
Enable metrics message production.
The default value is
Set the interval (in seconds) at which metrics messages are generated.
The default is 60 seconds.
Specify whether you want metrics messages to be persistent (that is, whether they will survive a broker failure).
The default is
Specify how long you want metrics messages to remain in their respective destinations before being deleted.
The default value is 300 seconds.
Set any access control you desire on metrics topic destinations.
See the discussion in Security and Access Considerations below.
Start up your metrics monitoring client.
When consumers subscribe to a metrics topic, the metrics topic destination will automatically be created. Once a metrics topic has been created, the broker's metrics message producer will begin sending metrics messages to the metrics topic.
There are two reasons to restrict access to metrics topic destinations:
Metrics data might include sensitive information about a broker and its resources.
Excessive numbers of subscriptions to metrics topic destinations might increase broker overhead and negatively affect performance.
Because of these considerations, it is advisable to restrict access to metrics topic destinations.
Monitoring clients are subject to the same authentication and authorization control as any other client. Only users maintained in the Message Queue user repository are allowed to connect to the broker.
You can provide additional protections by restricting access to specific metrics topic destinations through an access control file, as described in User Authorization.
For example, the following entries in an
accesscontrol.properties file will deny access to the mq.metrics.broker metrics topic to everyone except user1 and user 2.
The following entries will only allow users user3 to monitor topic t1.
Depending on the sensitivity of metrics data, you can also connect your metrics monitoring client to a broker using an encrypted connection. For information on using encrypted connections, see Message Encryption.
The metrics data outputs you get using the message-based monitoring API is a function of the metrics monitoring client you write. You are limited only by the data provided by the metrics generator in the broker. For a complete list of this data, see Metrics Information Reference.