How to determine a healthy agent?


Article ID: 181556


Updated On:


Management Platform (Formerly known as Notification Server)





How to determine a healthy agent?


·      Run through these healthy milestone checks:

·      Console’s Resource Explorer – Use the NS Console’s Resource Manager to determine if the Notification Server has a database record for that machine – (KB 19725) http://<NSServername>/Altiris/Resource/ResourceManagerConsole.aspx?Name=<machine name>

·      Is the client on the wire - Can you ping the machine? If the machines will not respond to a ping then you can’t expect successful communication between a client’s agent and the NS.

·      Does the resource exist in the NS - Once a machine’s agent is installed it will attempt to connect to its NS. If a connection is successful the NS will generate a unique guid and pass it to the client. The client will store that guid in the registry and in memory and send in Basic Inventory. That guid will be the unique identifier to the NS and the associated tables in the database. You can determine if the resource exists by using the Resource Manager in the Console or several tables in the database – Item, vItem, vResource, vResourceEx

·      Is the client ‘active’- The best method is the client’s Last Configuration Request (LCR). A resource which is not talking can and should be excluded from any health remediation, collections for Inventory Forwarding rules, collections for scopes and parent collections, and reports which determine compliances (example: Patch Management Solution). Using LCR for these basis need to Join on the Evt_NS_Client_Config_Request table where max(eventtime) > getdate() – X days. KB 27042. For hierarchal environments, LCR are events and events are not able to be forwarded until NS 7. Customers should use a use a DTS job from the reporting or parent NS to gather the LCR for all resources and store them in a new table. Then it can be referenced by your queries on your top NS. Or you can use MSSQL’s Linked Server query protocol to use the OPENQUERY method to determine the LCR of a resource(s) from another NS in real-time. Use KB 40338.

·      Can the client talk to the NS - Can the client get to NS using standard protocol? Get the NS connection string by using the Agent UI if RDP’d or gather the NS DNS name or IP from the client logs. Try to ping the NS using the connection string from the client.

·      Is the client receiving Policies - A successful communication can be determined by the agents’ ability to ask and receive a client policy. The client’s policy is a locally stored XML file where the agent is installed usually C:\Program Files\Altiris\Altiris Agent\Client Policies and has a file name of the NS. Open up the XML and see that it is getting new policies and look at the time stamp of the XML file. An outdated file is a sign that the agent is not successfully communicating. If you can, run the http request locally on the client to see if it the request is accepted. Open IE and use the URL: http://servername/AeXNS/Postevent.asp or https if using SSL - Use KB 2016. One thing to try is to delete the client policy and run a config request to see if a new one is created in the directory.

·      Is the client sending inventory and events to NS - Clients send inventory and events to the NS using http in the form of NSE files. When a machine runs inventory and or has an event such as executing a software delivery package or someone logs onto the machine it stores the NSE in the directory  C:\Program Files\Altiris\Altiris Agent\Queue - a folder with the NS name. Check to make sure NSEs are not backed up in this folder. Having more than one folder means the NS was connected to a different NS at one time and the old NS folder can safely be removed.

·      Are the agents registered – Altiris gathers agent inventory in the registry node under “HKLM\SOFTWARE\Altiris\Altiris Agent\Plugin Objects\Agents\Altiris Software Update Agent” and if the install did not properly lay down those keys then Altiris sees the Agent as not being installed. Running a repair of the agent should resolve the issue.

·      What do the Altiris logs say – In <program files>\Altiris\Altiris Agent\Logs are the Altiris agent logs. Look for any severity 1 log messages and paste the error into the Altiris Knowledgebase to see what maybe going on with the agent. Register the diagnostics dll to have the ability to right-click on the agent icon in the system tray and choose logviewer –”Regsvr32 AeXAgentDiagnostics.dll”

·      What does the Event Viewer tell you about the health of the client – If the OS or Hardware is having issues then it can impact the ability for the Altiris Agent to complete its processes. For example a corrupt file in the Altiris Inventory directory can cause the Agent to continually restart.