Deployment troubleshooting

In the event that you cannot deploy policies or instrumentation to a particular node, use these troubleshooting instructions to try to fix the problem. If the solution below does not fix the problem, click the check box to display the next suggested solution.

If deployment still does not work, you should contact your HP representative for more assistance. To help troubleshoot the problem, supply the error output log for the agent on the node where deployment fails. Error output for the agent can be found:
- UNIX: var/opt/OV/log/OpC/opcerror
- Windows: %OvAgentDir%\log\OpC\opcerror
Ensure that a firewall configuration is not preventing the connection between server and agent (for example, standard installation for Red Hat 7.1 installs firewall). If you need more information about connection through a firewall, contact your HP representative for a copy of the HP Operations Manager for Windows Firewall Concepts and Configuration Guide.
The firewall configuration is correct, but deployment still fails.
Ensure that the agent ID for the managed node is identical on the node and on the server. Refer to Agent ID on server and node mismatch for instructions.
The agent ID matches, but deployment still fails.
Ensure that the name of the management server (that you checked in the opcinfo file) can be resolved on the managed node. Refer to Resolve the IP address of the management server for instructions to verify this.
If the server node name cannot be resolved add the following statement to the nodeinfo file:
OPC_RESOLVE_IP <ip address of server>

Refer to Node Info Policy Parameters for more information about this parameter.

The name of the management server can be resolved on the managed node, but deployment still fails.
Check the opcinfo file on the managed node to ensure that the name of the management server is correct.
- UNIX Systems: /opt/OV/bin/OpC/install/
- Windows Systems: %OvAgentDir%\bin\OpC\install
If the computer listed as OPC_MGMT_SERVER is not the management server for the node, change it, save the opcinfo file, stop the agent with the command opcagt -kill, start the agent with the command opcagt -start
The management server is correct in opcinfo, but deployment still fails.
Check to ensure that the control agent is running on the managed node by typing the command opcagt -status If not, restart the agent with the command: opcagt -start
The agent is running, but deployment still fails.
The initial node deployment or a redeploy all deployment job fails with one of the following error messages:
- (AGENTADAPTER) Error: OvEpItoAdapter: (MS207) RPC resolve binding failed for '' ''. ...
- AGENTADAPTER) Error: OvEpItoAdapter: (MS202) opc_comm_connect() failed for 'Marrow'. (null). (MS203) RPC server on 'Marrow' not registered.
- (AGENTADAPTER) Error: OvEpItoAdapter: (MS208) RPC call to get character set from '' failed. RPC service 'Control Agent' on system '' returns an error: The RPC server is too busy to complete this operation.
The reason for these problems is that the installation of the agent software takes so long that the retries for the policy deployment time out. To solve this problem right-click Operations Manager in the console tree, and then click Configure Server.... In Namespaces, select Policy Management and Deployment, and then set the following values:
- Connect to managed node retry, which specifies the number of connection attempts (the default is 3)
- Time between connection attempts, which determines the delay in seconds between the individual connection attempts (the default is 5 seconds).
On slow machines the default values will not suffice and need to be increased accordingly.

All conditions appear to be met, but deployment still fails.
Check to ensure that the managed node meets all hardware and software requirements. Update the node as necessary before continuing. The requirements for all nodes can be found in the HP Operations Manager for Windows Installation Guide.
The hardware and software requirements are complete, but deployment still fails.
If a deployment job fails with an error message like "Unable to check Agent registry key..." or "Unable to connect to registry..." it means that the policy management server cannot access the registry of the remote managed node which is necessary to deploy packages.
Besides checking your network connection, you should also verify whether the policy management server has administrative rights on the remote managed node. Administrative rights are provided for the policy management server in a Windows domain environment by adding the domain group HP-OVE-Group to the local administrator group of the managed node.

If the managed node is not in the domain, but only in a Windows workgroup, administrative rights are provided by creating the local user HP-OVE-User on the remote managed node. This account must have the same password as the HP-OVE_User account specified during management server installation (see passthrough authentication.)

Additionally, the user HP-OVE-User must also be part of the local Administrator group of the managed node. If you have specified another user account for the server during product installation, you should create this account instead of HP-OVE-User on the managed node.

The network connections and administrative rights are correct, but deployment still fails.
If you are deploying a package to a Windows node, make sure that there are no errors reported in the agent installation log file. Open the file inst.log in \Temp or <install_dir>\Installed Packages\Temp on the managed node.
The inst.log file contains no errors, but deployment still fails.
If the managed node runs a Windows operating system, make sure that there are no errors reported in the Windows event log. To do this, you can use Event Viewer, which is provided with Windows.
The Windows Event Log on the managed node does not contain errors that would affect deployment, but deployment still fails.
If the error message, "Error granting privileges" appears in the Windows Node Security Setup window when attempting to add a node, it means that the node security setup failed. This can happen if you don't have administration rights for the node that you want to add.
You can gain administrative rights for the remote node in the following ways:
- In a domain environment, add the PMAD user (called HP-OVE-Deleg-User by default) to the Local Administrators group of the remote node.
- In a workgroup environment, you create a local account on the remote managed node. The account must have the identical name and password as the PMAD user account specified during product installation. Add this newly added account into the Local Administrators group.
The account permissions are correct, but deployment still fails.
Be sure that the node is not offline. Open a command prompt and use the ping command to see if the node responds.
The node responds to the ping command, but deployment still fails.
Starting with OVO 7.20, the Windows node Security Setup dialog box is no longer called when you move a managed node from the folder "Unmanaged Nodes with agent" from the left side to the right side of the Node Configuration editor after you have manually installed the agent on the node. Because the security dialog is not called in this case, the PMAD user (called HP-OVE-Deleg-User by default) is not automatically added to the local administrator group of the remote Windows node.
However, this user must be a part of the local administrators group of the remote node if you want to trigger any package deployment operation (deployment, undeployment, Redeploy all, or Uninstall) using the PMAD user. You will get a deployment error in these circumstances:
1. Install the Windows agent manually.
2. Add the node to the managed environment by moving it from the folder "Unmanaged nodes with agent" on the left side to the right side of the Node Configuration editor, then clicking OK.
3. Start any package deployment or undeployment operation.
To correct this situation, you can add the PMAD user manually to the local administrators group of the remote managed node of a domain environment. Any packaged deployment operation should work then, providing the remote node is not behind a firewall. For more details, see Start Windows node security setup.

Unmanaged node brought under management, but deployment still fails.
If the deployment job fails with an "Access Denied" and you chose the impersonation option, double-check that the fully qualified domain name (FQDN) is correct in the node's network properties.
Delegation mechanisms depend on end-to-end Kerberos authentication from the remote console, through the management server, and onwards to the node. This may only work correctly if the FQDN of the node is correct in the node's network properties. The FQDN in the node's network properties must be exactly the same as the node's full computer name, which you can configure in the system properties of the Windows operating system.

The FQDN must also be configured correctly in the dnsHostName and servicePrincipalName attributes of the computer object in Active Directory.

Start the command nslookup <shortname of agentsystem> on the node. This command should return the same FQDN as the full computer name in the node's Windows system properties.

If you are running a disjoint namespace, your nodes may have different FQDN domain suffixes, and the FQDNs may not all be registered correctly on the computer accounts. For more details about disjoint namespaces, see the following Microsoft Knowledge Base Article http://support.microsoft.com/default.aspx?scid=kb;EN-US;258503

If the FQDNs differ, you have to change your computer configuration so that DNS suffixes match in the Active Directory domain and on the node, or correct the DNS records to be present and computer account attributes.

The deployment job does not fail with an "Access Denied" error and the FQDN is correct in the node properties.
Check the node properties of the node where deployment fails to ensure that they are correct.
- System properties; are they correct for the node?
- Primary Node Name; is it set correctly?
- In case that the node name cannot be resolved (for example, due to missing DNS entry of the managed node): specify IP address directly.
- See Configure network information for managed nodes for more information about setting these parameters.
The network information for the node is correct, but deployment still fails.
If the node was recently configured, wait a few minutes until the security changes take effect, then restart the job.
Several minutes have passed since the node was configured, but deployment still fails.