لطفا منتظر باشید ...
3.2 Grid Control
The web-based Grid Control tool is a key part of Oracle Enterprise Manager
10g. You can use this tool to manage and monitor
multiple application servers through a single interface. This
interface provides a single view of all the application servers along
with links to each server's Application Server
Control. Because other infrastructure components are also monitored,
Grid Control gives you a way to measure and assure levels of
application service that leverage multiple components.While Application Server Control is used for administrative tasks on
an individual application server or cluster, Grid Control can manage
multiple application servers remotely and/or through firewalls.
|
3.2.1 Grid Control Infrastructure
The Grid Control infrastructure includes the following components
used to manage Oracle Application Server:Grid Control Console
The console, sometimes called the central
console, allows you to view, monitor, and manage
application servers as well as other Oracle components.
Oracle Management Service
Oracle Management Service, a J2EE application deployed on Oracle
Application Server, renders Oracle Enterprise
Manager's HTML user interface.
Management Repository
This central repository of enterprise-wide management data includes
many types of datafor example, hardware and software
configuration data leveraged in life-cycle, cloning, and patch
management. It also stores historic performance and availability data
used for trend analysis and reporting.
Oracle Management Agents
These agents monitor the application servers and communicate the
results of this monitoring back to the Oracle Management Service.
They also are responsible for running jobs for the Application Server
Control tool. You deploy one Management Agent on each host. The agent
will then monitor all components on that host, including all
installed Oracle software and non-Oracle components using Oracle
Enterprise Manager 10g's
extensibility.
The Grid Control architecture for managing multiple application
server targets is shown in Figure 3-2. The protocol used may be
either HTTP or HTTPS.
Figure 3-2. Grid Control architecture
systems on which Oracle Application Server runs and are responsible
for automatic service discovery, performance monitoring, and job
execution. Management Agents can also send Simple Network Management
Protocol (SNMP) traps to performance monitors in tools such as CA
Unicenter Network and Systems Management, the HP OpenView Operations
console, and IBM Tivoli. Agents can also perform any generic response
action, such as writing to a log file or another
console's API (which allows Oracle Enterprise
Manager to integrate well with other management tools and
procedures).
3.2.2 Using Grid Control
This section takes a closer look at managing Oracle Application
Server through Grid
Control. Logging into the console brings you to the console home
page. That page shows the status of all components in your
environment, as well as rollups of jobs, events, critical patches,
and deployments.Figure 3-3 shows Grid Control's
monitoring of an Oracle Application Server infrastructure. The
pull-down Deployments Summary selected in this figure lists all the
Oracle Application Server instances installed. You can also choose to
view only the status of application servers by choosing only
Application Servers under the View dropdown list on the upper left.
Figure 3-3. Oracle Enterprise Manager 10g's Grid Control console home page
selecting Application Servers under Target Search. This results in a
list of all application server targets, including high-level
statistics and status for each. Selecting a link to a particular
Oracle Application Server instance brings you to the Grid Control
Application Server home page for that instance (see Figure 3-4 for the top portion of how this page typically
appears).The Grid Control home page operates in a different way from the
Application Server Control home page, which you can access via the
Administer link in the Related Links section.Four links are shown, enabling you to quickly navigate the interface:
Home (your initial location), J2EE Applications, Web Applications,
and Performance.At the top of the console page, you will also find links to Setup
(for setting up and managing additional administrators, notification
methods, etc.), Preferences (for example, notification schedules),
Help, and Logout.
Figure 3-4. A portion of Grid Control's Application Server Control home page
This page is designed to give you a quick
view of the status of an application server and is segmented into the
following areas: general characteristics (e.g., status, availability,
URL, version, installation type, Oracle home, and host links), an
application URL response chart, a list of components (e.g., Oracle AS
Web Cache, Oracle HTTP Server), a list of current
alerts (for parameters such
as memory usage), a list of host alerts (for parameters such as CPU
utilization), and related links. The Host link takes you to the host
home page where you'll find performance and
configuration information for the host, as well as a software
inventory. Related links allow you to access other areas, such as the
patch list, metrics, manage metrics (with thresholds), alert history,
blackout history, monitoring configuration (if you have the OPERATOR
TARGET privilege), and administration (a login prompt for the
Application Server Control tool will appear). Of course, many of
these areas also include additional drill-downs.
J2EE Applications page
This page is typically used
to determine where J2EE application performance bottlenecks
are. It displays a list of J2EE applications, along with their total
processing time, average servlet/JSP processing time, number of
servlets/JSPs processed, total servlet/JSP processing time, average
EJB method execution time, number of EJB methods executed, and total
EJB method execution time. Because the screen shows resource usage,
you can drill down to find the source of bottlenecks.
Web Applications page
This page provides
an overview of all monitored web applications and their status, as
well as a summary of all outstanding
availability and
transaction performance alerts.
Drill-downs provide ASLM tools that enable you to ensure
high availability and optimal
performance of
web-based applications, as well as diagnose application performance
problems. The availability of web applications is defined by
administrator-specified criteria that determine whether your
application is up. Application Service Level Management tools also
let you monitor application performance through a multidimensional
approach: using synthetic transaction performance monitoring, and
measuring the actual application response times as experienced by
your real end users. Once a performance problem is detected, you can
interactively trace transactions from the client through the middle
tier down to the SQL statement level, thus enabling you to quickly
identify the components that may have contributed to application
bottlenecks. Historically, page performance is also traced through
the J2EE activity tiers, providing deep diagnostics that identify the
processing time consumed at the JSP, servlet, EJB method, and SQL
statement levels.
Performance page
This page graphically shows charts of
parameters such as CPU usage, memory usage, OracleAS Web Cache hit
rate, HTTP response and load, HTTP server process usage, active HTTP
connections, servlet response and load, and number of active
sessions. Refreshes can be manual or automatic at various intervals.
Historical trending of various time lengths can be viewed to trends
of up to 31 days. Through various links you can identify top servlets
and top JSPs (servlets and JSPs can be ranked by requests processed,
average processing time, or total processing time).
