Hi,

I am running my fluent job in batch mode on two linux parallel processors with a UDF. The UDF was compiled in parallel and i have been able to run for a long time without any problems. But yesterday, i got this error message:

999999: mpt_read: e999999: mpt_read: error:

I am not sure what is wrong but any help will be appreciated.

James

This problem is from the crash of Fluent node processes. Make sure no abnomality happens on the cluster during your iterations.

Hi RJ, This was indeed the problem. We had a problem with air conditioning and it led to node failures.

Thanks.

James