Oracle® Communications Service Broker Netra 6000 High Availability Manager Administrator's Guide Release 5.0 Part Number E20234-01 |
|
|
View PDF |
This chapter describes how to manage and monitor hardware and processes for Oracle Communications Service Broker Netra 6000 High Availability Manager (HA Manager).
Process management relates to the life cycle management of:
Service Broker Signaling Servers
Service Broker Processing Servers
SS7 SIGTRAN processes
Web Administration Console Servers
Logging processes
Process management controls the life cycle of Signaling Servers and Processing Servers. The servers can have the following states:
Shut down
Safe mode
Admin
Running
For a description of the life cycle states and transitions, see Oracle Communications Service Broker System Administrator's Guide.
Process management includes functions for:
Starting a process
Stopping a process
Terminating a process
The term process refers to:
Signaling Servers
Processing Servers
Logging Server
SS7 SIGTRAN
Web Administration Console Servers
When you restart a Signaling or Processing Server, it is gracefully shut down according to the HA Manager life cycle, started, and brought to the state it was in prior to the restart.
When you stop or terminate a process that is not a Signaling or Processing Server, the process state is managed at the operating system level.
Process management is handled by two components:
external Management Controller (eMC)
external Management Agent (eMA)
The eMC controls all Signaling Servers and Processing Servers in a deployment and spans multiple Service Broker domains.
eMA runs as a separate process on the same Worker Blade as the process or processes it controls. The eMA process is automatically started and stopped as a part of the operating system boot and shut down process. Only one eMA process runs on each Worker Blade.
A combined eMC and eMA process runs on the Bootstrap Blade. You use the Administration Console GUI to control this process.
You can monitor the state of the different processes and the blades using the Administration Console. The states are color coded as described in Table 5-1.
Table 5-1 Color Codes for Monitored Components
Color | Monitored Component | Description |
---|---|---|
Grey |
Signaling Servers Processing Servers Logging Server SS7 SIGTRAN stack Administration Console Web Servers |
The eMC in the Administration Console is disconnected from the eMA on the blade. The state is unknown. |
Grey |
Blade |
The blade in the slot is not monitored. The state is unknown. |
Black |
Signaling Servers Processing Servers |
The Signaling Server or Processing Server is in the Shut down state. |
Black |
Logging Server SS7 stack Administration Console Web Servers |
The process is not running. |
Black |
Blade |
The blade in the slot is powered off. |
Red |
Signaling Servers Processing Servers Logging Server SS7 stack Administration Console Web Servers |
The process is in the Error state. |
Orange |
Signaling Servers Processing Servers |
The process is in the Safe state. |
Blue |
Signaling Servers Processing Servers |
The process is in the Admin state. |
Green |
Signaling Servers Processing Servers |
The process is in the Running state. |
Green |
Logging Server SS7 stack Administration Console Web Servers |
The process is running. |
Green |
Blade |
The blade in the slot is powered on. |
Metrics available for Worker Blades are:
CPU utilization (percentage)
Number of HA Manager sessions handled
Metrics available for Signaling Servers and Process Servers are:
CPU utilization (percentage)
Heap Usage (MB)
Size of the Work Manager thread pool
Signaling Servers also provide the number of processed HA Manager events per second as a metric.
HA Manager Signaling Servers and Processing Servers are distributed over the available blades in a chassis. See Table 5-2 for a description of this distribution.
The servers are named using the following convention:
Signaling Servers: managed_ssu_chassisID_slotID_processID
Processing Servers: managed_pn_chassisID_slotID_processID
Where:
chassisID is the ID of the chassis
slotID is the ID of the slot within the chassis
processID is the unique identifier for the process running on the blade inserted in the slot
Table 5-2 Process Distribution in a single Chassis Deployment
Slots | Type | Processes |
---|---|---|
0 and 1 |
Bootstrap Blade |
Signaling Servers Administration Console Web Server (1) Processing Servers Administration Console Web Server (1) System Administration Console Web Server (1) Logging Server (1) |
2, 3, 4, 5 |
Worker Blade |
SS7 SIGTRAN (1) Signaling Servers (1) Processing Servers (2) |
6, 7, 8, 9 |
Worker Blade |
Processing Servers (3) |
In a multi-chassis setup, there is only one Bootstrap Blade in each chassis. The Bootstrap Blade is inserted in to slot 0. The Worker Blades are inserted into slots 1 through 9. This means that there is one more Worker Blade per chassis in a multi-chassis setup than there is in a single-chassis setup. This additional Worker Blade runs three Processing Servers.
The state of each process and blade is indicated by its color and corresponding text in the Administration Console. See Table 5-1, "Color Codes for Monitored Components".
The state of the blades is shown in the System tab Overview subtab.
The state for blades and processes are visible in the following subtab:
Chassis n
n is the ID of the chassis
Comparison_View
Comparison_View represents the name of a comparison view that you created. See "Comparing Blades" and "Comparing Signaling Server Processes and Processing Server Processes" for more information.
By default, the deployment is not monitored and all monitored-component icons in the console are grey. See Table 5-1, "Color Codes for Monitored Components". Statistics continue to be collected even when monitoring is stopped.
To start monitoring, click the Start Monitoring button:
To stop monitoring, click the Stop Monitoring button:
To set the metrics to monitor:
In the System Administration Console, click the System tab.
On the top right corner of the window, click the Settings icon (the wrench).
The Settings dialog box opens.
In the Slot list, select one of the following values to set the metric to monitor for the blades:
CPU: To monitor the CPU utilization on the blade
Sessions: To monitor the number of HA Manager sessions handled by the blade
In the Signaling Server list, select one of the following values to set the metric to monitor for a Signaling Server:
CPU: To monitor the process's CPU utilization
Sessions: To monitor the number of HA Manager sessions handled by the process
Heap Usage: To monitor the heap usage of the process
Event Rate: To monitor the number of Service Broker events per second that is handled by the process
In the Processing Server list, select one of the following values to set the metric to monitor for a Processing Server:
CPU: To monitor the process's CPU utilization
Sessions: To monitor the number of HA Manager sessions handled by the process
Heap Usage: To monitor the heap usage of the process
In the Refresh speed list, select one of the following values to set how often the statistics are refreshed in the GUI:
Slow: Every 3 seconds
Normal: Every 1.5 seconds
Fast: Every 0.5 second
Click OK.
To monitor a Worker Blade:
In the System Administration Console, click the System tab and then the Overview tab.
Locate the slot where the blade is inserted.
The metric is displayed as a bar graph and in numeric format. A graph showing historical data is also displayed.
To monitor the CPU utilization of a Signaling Server or Processing Server:
In the System Administration Console, click the System tab and then the Chassis n tab.
Locate the slot where the blade is inserted.
The metric is displayed as a bar graph and in numeric format. A graph showing historical data is also displayed.
To compare the metrics of set of blades:
In the System Administration Console, click the System tab and then the Overview tab.
The Overview tab for deployment is displayed.
Select the slots in which the blades you want to compare are inserted while holding down the Shift key.
Selected slots are outlined in yellow.
Right-click the selected slots.
A context menu opens.
Select Compare from the menu.
The Enter Name dialog box appears.
Enter a name for the comparison view and click OK.
A new tab is added for the comparison view. The tab is labeled with the name you entered. This tab is referred to as Comparison_View.
The tab displays the metrics for each slot.
To compare a set of Signaling Servers or a set of Processing Servers:
In the System Administration Console, click the System tab and then the Chassis tab of the chassis you want to monitor.
Select the server processes to compare while holding down the Shift key.
Selected server processes are outlined in yellow.
Right-click the selected processes.
A context menu for process management opens.
Select Compare from the menu.
The Enter Name dialog box appears.
Enter the name of the comparison view and click OK.
A new tab is added for the comparison view. The tab is labeled with the name you entered. This tab is referred to as Comparison_View.
The tab displays the process for each slot.
To change the state of a Processing Server or a Signaling Server:
In the System Administration Console, click the System tab and then the Chassis tab of the chassis you want to monitor.
The tab for the corresponding chassis is displayed.
Click the server process to select it. Select multiple server processes while holding down the Shift key.
Selected server processes are outlined in yellow.
Right-click the processes.
A context menu for process management opens.
Select one of the following options from the menu:
Go to Running
Transitions the server process to Running state.
Go to Admin
Transitions the server process to Admin state.
Go to Safe
Transitions the server process to Safe state.
Shutdown
Gracefully transitions the server process to Shutdown state.
Force Shutdown
Forces the server process to Shutdown state.
Restart
Gracefully shuts down the process according to the HA Manager life cycle, restarts it at the operating system level using the UNIX kill
command, and then transitions it to the state it was in before it was restarted.
Terminate
Terminates the process on an operating system level using the UNIX kill -9
command.
Compare
Compares two or more processes. See "Comparing Signaling Server Processes and Processing Server Processes" for more information.
To change the state of an Administration Console Web Server, Logging Server, or SS7 SIGTRAN process:
In the System Administration Console, click the System tab and then the Chassis tab of the chassis you want to monitor.
Click the process to select it. Select multiple processes while holding down the Shift key.
Selected processes are outlined in yellow.
Right-click the processes.
A context menu for process management opens.
Select one of the following options from the menu:
Start
Starts the selected processes.
Stop
Stops the selected processes at the operating system level using the UNIX kill
command.
Terminate
Terminates the selected processes at the operating system level using the UNIX kill -9
command.
To change the state of a blade:
In the System Administration Console, click the System tab and then the Overview tab.
The Overview tab for the deployment is displayed.
Click the slot in which the blade you want to change the state for is inserted. Select multiple slots while holding down the Shift key.
Selected slots are outlined in yellow.
Right-click the slots.
A context menu opens.
Choose one of the following options from the menu:
Power On
Powers on the blade.
Power Off
Powers off the blade.
Reset
Resets the blade by first powering it off and then on.