# When use MPI on two HPC nodes, receive the following error and program will exit
**URL:** https://discourse.paraview.org/t/when-use-mpi-on-two-hpc-nodes-receive-the-following-error-and-program-will-exit/12605
**Category:** ParaView Support
**Tags:** mpi
**Created:** [July 31, 2023, 9:28am UTC](https://discourse.paraview.org/t/when-use-mpi-on-two-hpc-nodes-receive-the-following-error-and-program-will-exit/12605 "2023-07-31T09:28:00Z")
**Posts on this page:** 10
**Page:** 1
### Author:  [@mahb](https://discourse.paraview.org/u/mahb)
#### Post date: [July 31, 2023, 9:28am UTC](https://discourse.paraview.org/t/when-use-mpi-on-two-hpc-nodes-receive-the-following-error-and-program-will-exit/12605/1 "2023-07-31T09:28:00Z")
use paraview 5.9.0,When use MPI on two HPC nodes, receive the following error. How to solve it ？ thanks
```auto
^[[0m^[[31m(7745.250s) [pvserver.0]vtkSocketCommunicator.c:781 ERR| vtkSocketCommunicator (0x1c09c10): Could not receive tag. 41235^[[0m
^[[0m^[[31m(7745.251s) [pvserver.0]vtkSocketCommunicator.c:781 ERR| vtkSocketCommunicator (0x1c09c10): Could not receive tag. 41233^[[0m
^[[0m^[[31m(7745.251s) [pvserver.0]vtkSocketCommunicator.c:781 ERR| vtkSocketCommunicator (0x1c09c10): Could not receive tag. 1^[[0m
^[[0m^[[31m(7745.251s) [pvserver.0]vtkTCPNetworkAccessMana:297 ERR| vtkTCPNetworkAccessManager (0x1b44260): Some error in socket processing.^[[0m
```
---
### Author:  [@mwestphal](https://discourse.paraview.org/u/mwestphal)
#### Post date: [July 31, 2023, 9:47am UTC](https://discourse.paraview.org/t/when-use-mpi-on-two-hpc-nodes-receive-the-following-error-and-program-will-exit/12605/2 "2023-07-31T09:47:30Z")
How did you install ParaVIew on the HPC ?
How did you install ParaVIew on the client ?
---
### Author:  [@utkarsh.ayachit](https://discourse.paraview.org/u/utkarsh.ayachit)
#### Post date: [July 31, 2023, 2:56pm UTC](https://discourse.paraview.org/t/when-use-mpi-on-two-hpc-nodes-receive-the-following-error-and-program-will-exit/12605/3 "2023-07-31T14:56:55Z")
Looking at the error, looks like something messed up when creating a view. Are there any other errors posted on the server side? That should illuminate the problem, Try running pvserver as follows to save output to a log file per process.
```bash
> pvserver -l=[some path]/pvserver.log,9
```
This will produce `pvserver.0.log, pvserver.1.log,...` log files that we can inspect to see what could be the problem.
---
### Author:  [@mahb](https://discourse.paraview.org/u/mahb)
#### Post date: [August 1, 2023, 7:40am UTC](https://discourse.paraview.org/t/when-use-mpi-on-two-hpc-nodes-receive-the-following-error-and-program-will-exit/12605/4 "2023-08-01T07:40:16Z")
I install paraview client on windows ;
start pvserver on HPC, by the following commands:
/home/mahb01/XXX/bin/mpirun -hosts gpu1,gpu2 -np 2 /XXX/bin/pvserver --force-offscreen-rendering --server-port=$SERVER\_PORT -l=“log32.txt,9” --displays=:0
---
### Author:  [@mwestphal](https://discourse.paraview.org/u/mwestphal)
#### Post date: [August 1, 2023, 7:45am UTC](https://discourse.paraview.org/t/when-use-mpi-on-two-hpc-nodes-receive-the-following-error-and-program-will-exit/12605/5 "2023-08-01T07:45:19Z")
> [@mahb](#):
>
> I install paraview client on windows ;
Ok but how ? Which version ?
And how did you install on the HPC ?
---
### Author:  [@mahb](https://discourse.paraview.org/u/mahb)
#### Post date: [August 1, 2023, 7:45am UTC](https://discourse.paraview.org/t/when-use-mpi-on-two-hpc-nodes-receive-the-following-error-and-program-will-exit/12605/6 "2023-08-01T07:45:46Z")
thanks. I get the log files but I just get the the following error message and can not find the problem.
(1315.990s) [pvserver.0]vtkSocketCommunicator.c:781 ERR| vtkSocketCommunicator (0x2570500): Could not receive tag. 1
(1315.990s) [pvserver.0]vtkTCPNetworkAccessMana:297 ERR| vtkTCPNetworkAccessManager (0x2325b60): Some error in socket processing.
(1316.067s) [pvserver.0] loguru.cpp:485 1| atexit
---
### Author:  [@mahb](https://discourse.paraview.org/u/mahb)
#### Post date: [August 1, 2023, 7:49am UTC](https://discourse.paraview.org/t/when-use-mpi-on-two-hpc-nodes-receive-the-following-error-and-program-will-exit/12605/7 "2023-08-01T07:49:43Z")
now the problem still exists.
and I have developed my plugin based on the version 5.9.0.
---
### Author:  [@utkarsh.ayachit](https://discourse.paraview.org/u/utkarsh.ayachit)
#### Post date: [August 1, 2023, 12:53pm UTC](https://discourse.paraview.org/t/when-use-mpi-on-two-hpc-nodes-receive-the-following-error-and-program-will-exit/12605/8 "2023-08-01T12:53:10Z")
Mind attaching the full contents. Often the real error is hidden in earlier in earlier messages. Also you should have two log files, one for each process. Let’s take a look at both to see if we can figure what’s going on.
---
### Author:  [@mahb](https://discourse.paraview.org/u/mahb)
#### Post date: [August 3, 2023, 6:13am UTC](https://discourse.paraview.org/t/when-use-mpi-on-two-hpc-nodes-receive-the-following-error-and-program-will-exit/12605/9 "2023-08-03T06:13:27Z")
[log31.txt.0](https://discourse.paraview.org/uploads/short-url/3yQluMKkxmgZmw9dmi2r6688YiB.0) (1.0 MB)
[log31.txt.1](https://discourse.paraview.org/uploads/short-url/97p0U9rQwRXeNAooDvneQhmiAwG.1) (1.0 MB)
Ok, I attach the logs . Help to see, thanks
---
### Author:  [@utkarsh.ayachit](https://discourse.paraview.org/u/utkarsh.ayachit)
#### Post date: [August 3, 2023, 6:10pm UTC](https://discourse.paraview.org/t/when-use-mpi-on-two-hpc-nodes-receive-the-following-error-and-program-will-exit/12605/10 "2023-08-03T18:10:30Z")
The log points to a possible issue with the `vtkNeuronInformationFilter`. It could be a bug in the filter itself or the reader upstream producing incorrect data or metadata, but in any case don’t think one (or both) of them is intended to work in parallel.