I was trying to run a job using cluster.
When I submit the job, error appeared as follows:
When submitting HPC jobs and running on the cluster, your analysis path must be a UNC path that all nodes can access. A UNC path is a path name that starts with \\
But it has no problem that set using same method for v2013.
Is there someone know how to solve?
To ask the obvious question is when you go to File-Open what is the path that you use to get to the file?
As the error mentions we would not want to use C:\somePath\Myfile, we would want to use \\MyComputerName\SharedFolder, or \\MyComputerName\c$\somePath\MyFile
Thanks for your reply. The path of the file is right.
I tested a same example, when submit a job from A to B is OK, and B to A is Ok.
But when using cluster to submit, error appeared: solver exits unexpectedly upon reaching the "processing input" stage.
Appreciate your help.
Thanks very much.
I have faced the same problem and Apollo's suggestion had solved it earlier.
There are two situations:
1) If you are submitting to cluster through your local machine, then it doesn't matter if you open the file through the local directory or UNC directory.
2) If you are using remote desktop to log in to the cluster's headnode, and submitting the jobs to cluster from there, you need to make sure you open the file through a UNC path and not the local path. That is, first you need to make sure your files are at the headnode's locl directory. Then you share it across the network. And then, open the file through this shared location, and not through the local directory.
Which one of these two situations are you mentioning?
I have tested to install SP2, but the error "Can not find a valid sequence of patches " appeared.
The error is same on both 32-bit and 64-bit, OS is Chinese.
But the SP1 has been fixed, I don't know why.