Skip to main content

FME server job for 23M records fails on one environment, but succeeds on another. Comparison of initial memory resources shows that the starting point for both jobs is the same:

log_comparison 

However, a comparison of memory fragmentation in those environments shows some differences as follow:

engine_comparison 

Is it possible that memory fragmentation causes that on one FME Server Engine job succeeded, but on the other failed?

Hi @zaneta.mark​ ,

 

It's possible memory fragmentation is the problem. It might be worth a) checking to make sure there are no other job/processes running on those machines when this job runs. and b) compare the swap/page file settings on the two machine to make sure they are using memory the same way.


I've faced memory issues in virtualised environments when the Windows VM has had memory allocated to it in the Hypervisor but not reserved. This was resulting in over allocation of memory on the host and causing things to fail on one machine, but not on the other


Reply