# Does paraview still work with CGNS files?
**URL:** https://discourse.paraview.org/t/does-paraview-still-work-with-cgns-files/704
**Category:** ParaView Support
**Created:** [October 10, 2018, 9:54pm UTC](https://discourse.paraview.org/t/does-paraview-still-work-with-cgns-files/704 "2018-10-10T21:54:03Z")
**Posts on this page:** 10
**Page:** 1
### Author:  [@albertzgd](https://discourse.paraview.org/u/albertzgd)
#### Post date: [October 10, 2018, 9:54pm UTC](https://discourse.paraview.org/t/does-paraview-still-work-with-cgns-files/704/1 "2018-10-10T21:54:03Z")
Hi all, I am a student working on computational fluid dynamics, and I do simulations which generates cgns files. I used to use paraview 4.2 to process the data and generate figures/videos, recently we updated our CGNS and HDF5 to a newer version, and paraview 4.2 stopped working with my input. And I found that newer versions of paraview doesn’t naturally support CGNS/HDF5 files.
Is there any sort of plugins which I can install in order to be able to process CGNS files with paraview?
Many Thanks!
---
### Author:  [@wascott](https://discourse.paraview.org/u/wascott)
#### Post date: [October 10, 2018, 11:19pm UTC](https://discourse.paraview.org/t/does-paraview-still-work-with-cgns-files/704/2 "2018-10-10T23:19:15Z")
Have you tried ParaView 5.5.2? Also, try a nightly build? There has been a lot of work in ParaView with CGNS over the last year or so.
If this doesn’t fix the issue, mind posting a dataset, and directions how to replicate the issue?
Alan
---
### Author:  [@albertzgd](https://discourse.paraview.org/u/albertzgd)
#### Post date: [October 11, 2018, 4:31pm UTC](https://discourse.paraview.org/t/does-paraview-still-work-with-cgns-files/704/3 "2018-10-11T16:31:32Z")
Hi, Thanks for your reply.
I tried using 5.5.2(I downloaded the binary, didn’t build from source code), the problem is like before, Paraview will give me a message saying that a reader for my file can’t be found and ask me to manually select a reader from a provided list where there is no CGNS reader.
My cgns files are built with parallel cgns version 3.3.1 and parallel hdf5 1.10.1, I can read and process them with python h5py module after it’s properly built from my local hdf5 library. Do I have to build paraview from source as well?
Thanks
---
### Author:  [@wascott](https://discourse.paraview.org/u/wascott)
#### Post date: [October 11, 2018, 5:31pm UTC](https://discourse.paraview.org/t/does-paraview-still-work-with-cgns-files/704/4 "2018-10-11T17:31:06Z")
Albert,
Could you post a file that shows this issue?
Alan
---
### Author:  [@albertzgd](https://discourse.paraview.org/u/albertzgd)
#### Post date: [October 11, 2018, 6:48pm UTC](https://discourse.paraview.org/t/does-paraview-still-work-with-cgns-files/704/5 "2018-10-11T18:48:12Z")
Alan,
I shared some of those files on google drive:
[https://drive.google.com/open?id=17ogl\_7HICficM61BvAPKIcCHEO\_45-PT](https://drive.google.com/open?id=17ogl_7HICficM61BvAPKIcCHEO_45-PT)
Thanks,
Albert
---
### Author:  [@wascott](https://discourse.paraview.org/u/wascott)
#### Post date: [October 11, 2018, 9:42pm UTC](https://discourse.paraview.org/t/does-paraview-still-work-with-cgns-files/704/6 "2018-10-11T21:42:29Z")
Current ParaView (5.5.2) does not recognize these files. Please download ParaView 5.6.0-rc2, from the ParaView website. This works correctly.
You gave me three files. Files flow-10.00.cgns opens correctly. grid.cgns opens correctly. part-10.00.cgns crashes. Could you try opening this dataset with cgnsview, and make sure your dataset isn’t bad?
---
### Author:  [@Joachim\_Hodara](https://discourse.paraview.org/u/Joachim_Hodara)
#### Post date: [May 19, 2020, 9:47am UTC](https://discourse.paraview.org/t/does-paraview-still-work-with-cgns-files/704/7 "2020-05-19T09:47:23Z")
Hi all,
I am experiencing a very similar issue using ParaView 5.8.0 on Windows. Everything works fine when reading small/medium size grids, but when I try to visualize large CGNS meshes (50M to 150M elements), ParaView can’t find an appropriate reader.
The official cgns checker passes, and I can visualize my grid using cgns view/plot.
Another interesting thing I noticed: On my Linux laptop (using ParaView 5.8.0), ParaView actually recognizes the CGNS format, but crashes later on because I don’t have enough RAM. However, a colleague tried to open a similar grid on Linux using ParaView 5.2 and faced the unrecognized reader issue. I will try installing a more recent version of ParaView on a Linux workstation to see if the issue was fixed in Linux between 5.2 and 5.8.
Is there something I am missing? What could I do to visualize my CGNS mesh in ParaView?
---
### Author:  [@jedbrown](https://discourse.paraview.org/u/jedbrown)
#### Post date: [April 30, 2022, 3:18am UTC](https://discourse.paraview.org/t/does-paraview-still-work-with-cgns-files/704/8 "2022-04-30T03:18:33Z")
I’m using Paraview 5.10.1 and it isn’t offering any fields in a simple file (generated by the CGNS user manual examples). This file has “Density” and “Pressure” fields when opened in VisIt, but no such fields in Paraview.
[grid\_c.cgns](https://discourse.paraview.org/uploads/short-url/40ElHUDkUeyk79n4HpG9hwtRl0G.cgns) (245.6 KB)
---
### Author:  [@utkarsh.ayachit](https://discourse.paraview.org/u/utkarsh.ayachit)
#### Post date: [May 2, 2022, 5:19pm UTC](https://discourse.paraview.org/t/does-paraview-still-work-with-cgns-files/704/9 "2022-05-02T17:19:01Z")
Make sure you check arrays listed under **Point Arrays**. With the `grid_c.cgns` file you attached, this is what I see. **Density** and **Pressure** are indeed available, just not read by default. You need to enable them in the reader’s **Properties**.
---
### Author:  [@jedbrown](https://discourse.paraview.org/u/jedbrown)
#### Post date: [May 2, 2022, 6:00pm UTC](https://discourse.paraview.org/t/does-paraview-still-work-with-cgns-files/704/10 "2022-05-02T18:00:16Z")
Oh, how embarrassing. Thanks, that’s working.
Is it expected to work with higher order elements (cubic 0036 and quartic 0038 [CFD General Notation System](https://cgns.github.io/Proposals.html))? I get this with the cubic elements produced by `src/tests/elemtest.c`, but `cgnscheck elemtest.cgns` has warnings too so I don’t know if it’s a fair test. I’d be happy if reading HEXA\_64 is expected to work (reducing to linear internally is fine).
```auto
( 6.500s) [paraview] vtkCGNSReader.cxx:2447 ERR| vtkCGNSReader (0x55821eb691b0): Mixed definition of unstructured zone by elements and by faces is not valid.
( 6.503s) [paraview] vtkCGNSReader.cxx:4608 ERR| vtkCGNSReader (0x55821eb691b0): Error Reading file
( 6.503s) [paraview] vtkExecutive.cxx:752 ERR| vtkPVCompositeDataPipeline (0x55821c2922c0): Algorithm vtkCGNSFileSeriesReader(0x7f4d340087e0) returned failure for request: vtkInformation (0x55821dda85d0)
Debug: Off
Modified Time: 407151
Reference Count: 1
Registered Events: (none)
Request: REQUEST_DATA
FROM_OUTPUT_PORT: 0
ALGORITHM_AFTER_FORWARD: 1
FORWARD_DIRECTION: 0
```
[elemtest.cgns](https://discourse.paraview.org/uploads/short-url/yKULo98ieTfasbEtJOZppHi4fed.cgns) (118.5 KB)