# Saving meshes in a list: RuntimeError: \*\*\* Error: Duplication of MPI communicator failed **URL:** \ **Category:** mesh\ **Created:** [February 28, 2022, 9:15am UTC](https://fenicsproject.discourse.group/t/saving-meshes-in-a-list-runtimeerror-error-duplication-of-mpi-communicator-failed/7749 "2022-02-28T09:15:05Z")\ **Posts on this page:** 1\ **Showing post:** 8
**Author:** ![dokken](https://yyz2.discourse-cdn.com/free1/user_avatar/fenicsproject.discourse.group/dokken/32/1560_2.png) [@dokken](https://fenicsproject.discourse.group/u/dokken)\ **Post date:** [March 4, 2025, 12:46pm UTC](https://fenicsproject.discourse.group/t/saving-meshes-in-a-list-runtimeerror-error-duplication-of-mpi-communicator-failed/7749/8 "2025-03-04T12:46:11Z")
> [@chenyongxin](#): > > I think I met a similar situation, where my code terminates at around 1,900 steps with the following error output. If I save checkpoint at 1,000 step and restart from that point, the simulation can pass it. I have no direction why it happens. From the error output, if it can be deduced to the MPI duplication problem? In my code, I did not have an explcit call of generating new mesh in every time step. > > `` Generating a mesh at each time step might be problematic, as MPI only allows for a limited number of `duplicate`-calls within each simulation. Without the actual code you are running, it is hard to give any guidance on how to avoid it. --- _[View the full topic](https://fenicsproject.discourse.group/t/saving-meshes-in-a-list-runtimeerror-error-duplication-of-mpi-communicator-failed/7749)._