When inconsistencies are detected between the nodes, VERITAS Cluster Server (VCS) may not be able to probe specific resources to determine their current state. If VCS is unable to determine the state of any resources in the group, the Engine will return Error 10162, indicating the group name and the node name where it could not complete the probe. (Figure 1)
Figure 1
Resolution:
While the error message is helpful in determining the group and the node having difficulty, it does not indicate which specific resource is having problems. To do this, follow the steps listed below:
1. Open a command prompt on the one of the clustered nodes and issue an "hares -display" command (Figure 2). Due to the size of the amount of information this command produces, it is recommended that the output be piped to a text file.
Figure 2
2. The output lists each clustered resource, its attribute values, and its current state on both nodes. Scan through the text file, looking specifically at the values indicated for the "probed" attribute. (Figure 3)
Figure 3
3. The example indicates that there is a problem with the SQL_MSSQL resource on node VCSTEST2 that needs administrative attention. The problems can range from not having the agent installed on the node (missing .DLL file) to incorrect values specified for the resource. Correct the problem on the node so that VCS is able to fully probe the resource. (To do this, it may be necessary to run HASTOP -local to correct the issue. Once the issue is corrected, HAD can be restarted by issuing hastart.)