I got a error to run cluster computing with HPC server 2008 R2. I set up one head nodea and one compute node with network topology 1.
I already check everything related to run cluster computing using clusrun command. It looks everything is fine but when I run batch calculation and I got the following error on job management.
"Unable to access working directory on node %computenode%"
I check the network mapping on compute node and the the head node server is detected. And the folder which is used by compute node to calculate and is on head node is shared and I can access this folder using compute node.
But I don't know what cause this error. On the node management two server stay as health and online. the clusrun command works perfectly.
Is there anything I missed? Please give me any comments about this problem. Thanks in advance.
Can you share your "working directory" setting in your failed task? If you didn't set, the default working directory is "C:\Users\<yourAccount>", please check from the compute node whether you have access to that folder.
And also, please check whether your job's runas account has the access to the share on the headnode if your job is runas different credential from your clusrun.