Determining the Status of the eG Agents
The eG manager is able to determine and report the operational status of all the eG agents in the target environment. The sections that follow will discuss how to view this status information.
If you select the Agent Status option from the Agents tile, you will be lead to Figure 1 , which will provide status information for agents based on the agent types.
To obtain the status of the eG agents of a particular type, the administrator has to first select the type of agent (whether basic, premium, external, or remote) from the Agent type list box. If the Named User license is enabled, the eG Enterprise system provides you with a single, central interface using which you can instantly ascertain the current status of the internal agents (if any) that are monitoring the thin client and VDI components in your environment! In this case, an additional Agents for user licensed types option will appear in the Agent type list box.
The IP address / host names of the agents of the selected type will then be displayed. A symbol against each agent indicates that the agent has been deployed. A symbol appears against each agent implying that the agent has not been deployed. While the symbol indicates that the agent is running currently, the Status column indicates that the eG agent is not running.
Also, using the Search text box, you can find out the status of a particular agent. To know the status of a particular agent, just specify the IP address / host name of that agent in the Search text box, and then click the ‘magnifying glass’ icon next to it. The status of the specified agent will then appear. If the exact IP address / host name of the agent is not known, then a string or a character that features in the IP / host name of the agent can be provided in this text box (see ). Multiple search conditions can be specified as a comma-separated list.
To know the agents that are currently in a particular state, simply select an Agent status (which can be Running/Not Running/All). The default selection here is All.
You can even remotely initiate an agent-restart, by simply clicking on the Restart icon that corresponds to an agent. To restart all agents, click on the Restart All Agents button in Figure 2. Doing so immediately sends out restart requests to all the agents that are currently running and reporting metrics to the eG manager. If an agent is not running currently, then the eG Enterprise system sends out the restart request soon after that agent starts running.
You can also stop an eG agent from the eG management console itself, using this page. Click on the Stop icon corresponding to an eG agent in Figure 1, to stop that agent. Doing so immediately sends out a stop request to that agent. The next time that agent sends metrics to the manager, it reads the stop request and serves it.
If an administrator needs to be alerted upon login, about agents that are not reporting measures to the manager, then do the following:
- Open the eg_services.ini file in the <EG_HOME_DIR>/manager/config directory.
- In the [misc_args] section, set the AlertAgentsNotRunning flag to Yes (default is No).
Once this is done, the next time the administrator logs into the admin interface, a message listing the agents that are not running will be displayed.
An eG agent can be configured to run specific tests once a day or once every few hours. You can configure the eG manager to exclude tests that are infrequently run when it determines whether an agent is running or not. To do this, modify the value of NotReportingCutoffFactor in the [MISC_ARGS] section of the eg_services.ini file. By default, tests running with measure period of greater than 20 minutes are not considered by the eG manager for determining if an agent is running or not.
Also, by default, output logging is disabled for the eG agents configured in an environment. The eG Enterprise system allows you to enable output and error logging for a specific agent from the eG administrative interface itself, thereby saving you the trouble of running the debugon.bat file to achieve the same. When output logging is enabled, an agentout.log file is created in the <EG_INSTALL_DIR>\agent\logs directory to which details of the tests run and measures reported by that agent to the manager are recorded. To enable output logging for an agent, set the Output Logging Enabled? flag for that agent to on. When you attempt to enable output logging, a message box shown by Figure 3 will appear, requesting your confirmation to enable output logging for that agent.
Click on the ok button in the message box to enable output logging or the Cancel button to disable it. You can then click on the logs icon that corresponds to an agent in to view both error logs and output logs related to that agent. Clicking on the logs icon corresponding to that agent will lead you to Figure 4, where the contents of the error_log of the corresponding agent can be viewed by default.
If you have turned on output logging for an eG agent using the agents – status page, then you should not turn off output logging for that eG agent by manually running the debugon.bat file. Likewise, if you have turned on output logging for an eG agent by running the debugon.bat file, then you should not turn it off using the agents – status page.
You can pick any log file from the Log file name list to view its contents (see Figure 5).
- At any given point in time, you can click the Refresh button at the right, top corner of the area where the log file contents are displayed to refresh the contents of the log file. This way, you can make sure that the log file you are viewing is up to date.
- You can also click on the Download button next to the Refresh button to download the chosen log file.
Clicking on an agent displayed in Figure 1 will lead the users to an Agent Information page (see), which provides some agent-related information. This includes:
- The Agent IP/Nickname
- An indicator as to whether the auto upgrade capability has been Enabled for that agent, or Disabled
- The ID of the last upgraded package (if any) (if no upgrading has occurred, then this will be ‘None”)
- The date and time at which the agent was last upgraded
- The HostName of the agent
- The operating system on which the agent is executing
- The current version of the agent
- The date and time at which the agent last updated the manager with configuration changes
- A Reset button
- A Restart Agent button
- Once an agent is upgraded, information regarding the upgraded package will be registered with the manager. Figure 6 provides that information. Now, the next time the agent requests for an upgrade, the manager checks whether any newer upgrades are available. If any such upgrade is found, it sends the same to the agent. If for some reason the information pertaining to the last upgrade has to be cleared from the agent’s upgrade history, then click on the Reset button. This ensures that the details of the last upgrade are lost, and helps the agent download the last upgrade once again from the manager. To restart the agent, click on the Restart Agent button in . To view agent logs, click on the View Agent Log Files button in .
Moreover, if the Agent type chosen from Figure 1 is External Agents or Remote Agents, then you will also be able to view the count of hosts assigned (if any) to each external/remote agent. For this, you will have to click on the ‘+’ button that pre-fixes an agent (see Figure 7). Beneath the assigned host count, you can see that Figure 19.35 also reveals which specific hosts have been assigned to that agent. From a single glance therefore, you can precisely identify the external/remote agents that are been actively utilized, and those that are not.
To add more hosts to an external/remote agent, click on the button corresponding to an agent in Figure 7. This will open Figure 8, using which you can assign more hosts to the agent or disassociate existing hosts from it.
If the Agent type chosen from Figure 1 is Agents for user licensed types, then you will be able to view the status (whether installed or not and whether running or not) of those agents that are monitoring the thin client and VDI components in your environment in an agent-based manner – in other words, by selecting this option, you can determine the status of those internal agents that are monitoring the thin client and VDI components.
Figure 9 : Viewing the status of the agents of type 'Agents for user license types'