# Complete compilation of Paraview (Readers+OSmesa+Python)

**URL:** https://discourse.paraview.org/t/complete-compilation-of-paraview-readers-osmesa-python/10558
**Category:** ParaView Support
**Created:** [October 20, 2022, 2:52pm UTC](https://discourse.paraview.org/t/complete-compilation-of-paraview-readers-osmesa-python/10558 "2022-10-20T14:52:40Z")
**Posts on this page:** 1
**Showing post:** 8

<div class="post-metadata">

### Author: ![Captain\_Convergence](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/c/e19b73/32.png) [@Captain\_Convergence](https://discourse.paraview.org/u/Captain_Convergence)
#### Post date: [October 25, 2022, 10:35am UTC](https://discourse.paraview.org/t/complete-compilation-of-paraview-readers-osmesa-python/10558/8 "2022-10-25T10:35:12Z")

</div>

The reason for this is that I need to use ParaView in parallel to compute some field in a reasonable time but I noticed that I have problems when using multiple CPUs, see my previous post here to now more about the issue:

> [@Parallel vs Serial post-procesing](https://discourse.paraview.org/t/parallel-vs-serial-post-procesing/8324):
>
> Dear support I am trying to generate animations in Paraview but I am facing problem when using parallel post-procesing. I have attached an image where you can observe the Q-crit obtained by using multiple CPUs (top image) and using a single CPU (bottom figure). The lines in Q-crit contour are the limits of the domain blocks coming from the CFD solver which imports an extra cell and this modifies the spatial gradients in the domain. It seems that for a single CPU it is handled in a better manner…

However, I ran by chance in our HPC the ParaView version with osmesa using multiple-CPUs and I didn’t have the problem so now I believe that the solution to all my issues is to use a compiled version of ParaView using osmesa since I normally do CPU-based rendering.

---

_[View the full topic](https://discourse.paraview.org/t/complete-compilation-of-paraview-readers-osmesa-python/10558)._