Overview
|
|
This topic describes how to use the various screens available on the
Service Monitor web-based interface. It assumes that you have already performed
the steps described in the Service Monitor
Installation Guide.
|
Topology Overview
|
|
To monitor the traffic between Enterprise Gateway instances and various Web Services,
Clients, and Remote Hosts, running throughout your network, you must add the
Enterprise Gateway instance as a node on the Topology
overview page. This page shows all nodes (Enterprise Gateway instances) that
are sending monitored traffic to protected Web Services, Clients, and Remote
Hosts.
Each of these Enterprise Gateway nodes must already be configured to store message metrics
in the same database that Service Monitor is configured to read from. This enables the
Real-time Monitoring, Reports, and Audit
Trail screens to obtain the Enterprise Gateway metrics and logs that they display
from this database.
On the Topology overview page, if the database on the Enterprise Gateway
server is configured correctly, a database icon with a green up arrow is displayed
on the Enterprise Gateway node. However, if the Enterprise Gateway is not configured correctly,
a red down arrow is displayed. For details on configuring database connections,
see the Service Monitor Installation Guide.
Each node on the Topology overview page shows the total number
of messages processed by the Enterprise Gateway that it represents. The message count
is updated every 5 seconds. If the Service Monitor can not connect to a node, a
connection error is displayed on the node.
Adding a New Node
To add a new node, click the Add node button at the bottom
of the page. You must configure the following fields when adding a new node:
Host URL:
Enter the URL of the Enterprise Gateway that is protecting traffic that you want
to monitor. The URL entered here should include the host name and port
number of the Enterprise Gateway, for example:
|
| | |
|
http://HostName:8090/metrics
| |
| | |
|
where HostName is the host name of the machine
on which the Enterprise Gateway is running, and port 8090 is
the port used by the Enterprise Gateway's management interface.
Custom Name:
Enter an optional friendly name for this node. This may be useful to
help distinguish between different instances of the Enterprise Gateway in cases
where you have deployed several Enterprise Gateways throughout your network.
Username:
If the management interface on the node you are connecting to requires authentication
(for example, HTTP Basic authentication), you must enter the username to authenticate
with in this field.
Password:
Enter the password for this user in the field provided.
On the Topology overview page, you can right-click a node, and view
reports for all nodes, real-time monitoring for that node, or the audit trail for that
node, using the Reports, Real-time Monitoring, and
Audit Trail options respectively. There are also buttons for these
options available on the bottom of the Topology overview page. The
following sections describe each of these options:
|
Reports
|
|
Service Monitor uses message metrics stored in a centralized database by the Enterprise Gateway
instances running throughout your network. The Enterprise Gateway stores metrics for the
Web Services that it exposes (virtualized services), and for the Web Services, and
Client and Remote Host connections that it protects. Service Monitor can then generate
usage reports and charts based on this data. The Reports screen
is responsible for rendering these charts. The following configuration options
apply to each of the three tabs available on this screen.
Per-process statistics:
Reports can be generated for monitored objects per-process or
per-cluster. Currently, all Enterprise Gateway processes in Service Monitor’s
database belong to a single cluster (domain). An identity makes
metrics aggregated from corresponding objects in Enterprise Gateway configurations available in
a cluster. For example, Web Services of the same name configured on Enterprise Gateway processes
in a cluster make up a single Web Service identity.
So if Service A is configured on host 1, and Service A on host 2 (both in the same cluster),
these make up the Service A identity. In reports with per-cluster statistics, the values for
service A are aggregated from statistics for service A on both hosts. When generating reports
with per-process statistics, Service A on host 1, and Service A on host 2 are taken as separate
entities, and separate metrics for each are available.
In Service Monitor, you can specify whether per-process or per-cluster statistics are generated using
the Per-process statistics checkbox, which is disabled by default. Depending on
this selection, the drop-down list of items to generate in the report shows individual Web Service
items on individual hosts or Web Service identities (one Web Service identity for all Web Services
under the same name configured on hosts in the cluster). When this setting is unselected, the
Remote hosts list displays cluster-wide Remote host identities. When selected,
the Remote hosts list displays individual per-process Remote hosts.
Aggregated Metrics:
Aggregated Metrics display metrics for a group of monitored objects (for example, all Web Services)
across one or more nodes over a specific time period. This report displays the time line using a
graph. This enables you to view the total performance metrics for all Web Services, Remote Hosts,
or Clients.
Totals:
Use this option to show all values for a specific monitored object (for example, a Web Service)
over a specific time period. The report uses a bar chart to display the total value of a single
metric for a monitored object over a specified time period. This enables you to compare the
performance metrics of specific Web Services, Remote Hosts, or Clients.
Time Window:
You must select the time window size that you want to generate reports for using the button at the
top of the screen. The selected option must match the time window specified
in the Real time monitoring settings dialog when configuring the real-time
monitoring settings for the Enterprise Gateway. For example, if you selected 5 minutes
for the Time window to store on the Real time monitoring settings
dialog, you must select the 5 minute time window metrics when generating reports
on the following tabs:
- Remote Hosts
- Web Services
- Clients
In addition, all monitored Enterprise Gateway nodes must use the same Time window to
store setting.
Remote Hosts:
The Remote Hosts drop-down list contains all Remote Hosts that have been configured on
Enterprise Gateways for which monitoring has been enabled. You can select one or more Remote
Hosts to produce a report for.
Aggregates reports for Remote Hosts show the following:
-
Response times in terms of the number of requests processed in the specified
response time ranges.
-
Remote Host uptime, which details the results of a Watchdog configured
for a Remote Host. You can add Watchdogs to Remote Hosts by right-clicking
the Remote Host node on Services tab in the Policy Studio,
and selecting Watchdog -> Add. For more details, see
the Watchdog Configuration
topic.
-
Total number of bytes transferred into and out of this Remote Host.
Totals reports for Remote Hosts display the following metrics:
-
Response times.
-
Bytes transferred.
Web Services:
The Web Services drop-down list includes all Web Services
that have been invoked on the Enterprise Gateway configured for monitoring. In
particular, the Set Web Service Context filter must have
been invoked. You can select one or more Web Services from the list to
generate reports on these Web Services.
Aggregates and Totals reports show the total number of messages that have
been processed by the Web Service, in terms of those that have been blocked,
those that have passed, and those that have failed.
Clients:
On this tab, the Web Services drop-down list displays
all Web Services that have been invoked and had their Set Web
Service Context filter called. Select one or more Web Services
from this list. The Clients drop-down list is then
populated with all clients that have been authenticated for the selected
Web Services.
Aggregates and Totals reports show message metrics for clients that have
accessed the selected Web Service. This enables administrators to view which
users are calling which Web Services.
|
Real-Time Monitoring
|
|
You must enable real-time monitoring on the Enterprise Gateway to monitor real-time usage
statistics on the Real-time monitoring screens.
Enabling Real-Time Monitoring
To enable real-time monitoring, perform the following steps:
- In the Policy Studio tree view, right-click the HTTP Services Group
(for example, Default Services), and select Edit.
- In the HTTP Services dialog, ensure the
Include in real-time monitoring checkbox is selected.
- Click OK.
To enable monitoring for the entire Enterprise Gateway process, perform the following steps:
- On the Services tab in the Policy Studio, right-click the
Oracle Enterprise Gateway process, and select Monitoring
-> Traffic.
- In the Traffic Monitor Settings dialog, ensure the
Enable Traffic Monitor setting is selected.
- Click OK.
- Right-click the Oracle Enterprise Gateway Process node,
and select Monitoring -> Metrics.
- In the Real time monitoring settings dialog, ensure the
Enable real time monitoring setting is selected.
- Click OK.
Important Note:
Enabling traffic monitoring has a negative impact on performance. If you wish to maximize
performance, do not enable these settings. For more details, see
Configuring Traffic Monitoring.
When you have configured the Enterprise Gateway for monitoring, you can monitor traffic secured
(or blocked) by the Enterprise Gateway using the Real-time monitoring screen.
The tabs on this screen are described as follows:
Status:
This tab shows the total number of messages processed by the system. There are
separate charts to illustrate graphically the number of messages that have passed,
that have been blocked, and that have failed.
The Latest HTTP Messages table on this tab shows traffic monitoring
statistics for all HTTP messages processed by the currently selected Enterprise Gateway. You
can click the Refresh button to update this list. For example, the
available details per message include the following:
| Time |
The time that the message was processed at.
|
| Status |
The HTTP status code of the message (for example,
200 for OK, or 403 for Forbidden).
|
| Method |
The HTTP method (for example, GET or POST).
|
| Local Address |
The local address on which the Enterprise Gateway is installed.
|
| URI |
The relative path that the message is sent to (for example,
axis/services/urn:cominfo).
|
| Remote Address |
The remote address on which the Web Service resides.
|
| Remote Name |
The remote hostname on which the Web Service resides.
|
The Real-time monitoring console also displays the message path and the contents of
each request message and response message. For example, to view the message path and the content of
the response message sent by a Web Service through the Enterprise Gateway, perform the following steps:
- In the Latest HTTP Messages list, select the message whose contents
you wish to view. This displays the message path for the selected message under the list
on the left.
- Select the outbound message from the Enterprise Gateway, which is second in the list on the
right of the message path.
- Click the View button for the Response
message to view its contents.
- Alternatively, click the Download button for the Response
message to download its contents and save them to a file.
Audit and Alert messages:
This tab displays tables for Audit messages,
Alert messages, and SLA breach messages.
Each table displays various details about the message, including time, message,
message ID, and the filter or source of the message.
Each table displays the last 50 messages for the category that it represents.
For example, the Audit messages table displays the last 50
log messages, while the SLA (Service Level Agreement) breach table shows the
last 50 SLA breaches that have been triggered throughout the system.
For more information on triggering alerts, setting SLAs, and configuring
custom logging, see Alerting and Logging.
Remote hosts:
This tab lists all remote hosts that have been configured on a particular node.
The Configured remote hosts table shows the number of messages
that have been sent to this remote host, together with the total number of bytes
sent to and received from this host.
You can view more detailed information about a particular remote host by
clicking it in the table. The Message details table shows
general performance metrics for the remote host, including total number of
requests, number of requests/second, bytes sent and received, and
transfer rate in bytes/second.
The Remote hosts response codes table shows the total number of
each class of HTTP response status that was returned by the remote host. Typically,
a remote host returns a 200 status code when it has processed the message correctly,
but it may return a 4xx or 5xx message to indicate that an error has occurred.
You can use this table to gauge whether the remote host is processing messages
successfully or rejecting them.
The Remote hosts response times table categorizes the response times
of the remote host in terms of certain pre-configured time intervals. Administrators
can easily examine how their remote hosts are responding to requests in this
table.
Finally, you can use the Remote host up/down table in conjunction
with Watchdogs that have been configured on a remote host to determine the availability
of the remote host. When a Watchdog is configured for a remote host, it polls the
remote hosts at regular intervals by sending requests to it, and examining the responses.
This table displays the total number of Watchdog requests that have been deemed a success,
and those that have been deemed a failure, depending on the criteria configured for the
Watchdog.
To add a Watchdog to a remote host, right-click the remote host node in the
Policy Studio tree, and select Watchdog -> Add. For more details,
see Configuring an HTTP Watchdog.
Web Services:
The Web Services tab shows usage statistics for services that are exposed
by a particular node. These Web Services are created and exposed by importing service
definitions from WSDL files. To do this, in the Policy Studio tree view, right-click
the Web Services node, and select Register Web Service.
Follow the steps in the wizard to expose one or more of the operations defined in the WSDL.
Alternatively, you can also register Web Services using the web-based Service Manager tool.
For details, see Managing Web Services.
Traffic for Web Services that have been exposed by the Enterprise Gateway in this manner can
be monitored on this tab. The Web service usage table lists all Web
Services that are currently exposed by the Enterprise Gateway. You can click a Web Service
in the table to display more detailed metrics for this service.
Note:
A Web Service must have been called before it is displayed in
the Web service usage table.
|
Audit Trail
|
|
The Audit Trail enables you to filter log messages generated by a
particular node. You can view the filters by clicking the
Advanced Search button. You can then filter log
messages by time period, severity level, filter type, and filter name.
You can even filter log messages based on the text of the log message
itself. When you have configured the relevant filters, you can click
the Run button to search all log messages that match
the specified criteria.
The log messages that match the search criteria specified in the filter
are displayed in the table. The details in the table include the
log level (severity), time, log message text, and message ID.
|
|