[BUG] vtkMEDReader integration

[krzysiek-git]

Hi.
There’s something wrong with vtkMEDReader integration in Paravis in Salome 9.14 Windows native.
When you close hdf file, reopen it, go to Paravis and reload first file in pipeline the results disappear from the viewport. Message window tells me that there’s no *.rmed file I was reloading. The file is on disk, here:

C:\Users\KrzysztofZurawski\FEA\002_ENTERFEA_NL\06_snap_back_arc-length
I can open it and see the results. The problem is vtkMEDReader is trying to reloade file from some other location, which is not the place the file results file was originally stored:

C:\Users\KrzysztofZurawski\AppData\Roaming\938230184_\06-snap_back.rmed
Info from Message Window:
``
Exception has been thrown in vtkMEDReader::RequestInformation : File : "C:\Users\KrzysztofZurawski\AppData\Roaming\938230184_\06-snap_back.rmed" has been detected as NOT EXISTING : impossible to read anything !

ERROR: In vtkExecutive.cxx, line 741 vtkPVCompositeDataPipeline (0000020160135690): Algorithm vtkFileSeriesGroupReader (00000201600C73D0) returned failure for request: vtkInformation (000002016CBF8710) Debug: Off Modified Time: 1694914 Reference Count: 1 Registered Events: (none) Request: REQUEST_INFORMATION FORWARD_DIRECTION: 0 ALGORITHM_AFTER_FORWARD: 1

Exception has been thrown in vtkMEDReader::RequestInformation : File : "C:\Users\KrzysztofZurawski\AppData\Roaming\938230184_\06-snap_back.rmed" has been detected as NOT EXISTING : impossible to read anything !

ERROR: In vtkExecutive.cxx, line 741 vtkPVCompositeDataPipeline (00000201601360B0): Algorithm vtkMEDReader (0000020160133950) returned failure for request: vtkInformation (000002016CBF9580) Debug: Off Modified Time: 1694945 Reference Count: 1 Registered Events: (none) Request: REQUEST_INFORMATION FORWARD_DIRECTION: 0 ALGORITHM_AFTER_FORWARD: 1

Exception has been thrown in vtkMEDReader::RequestInformation : File : "C:\Users\KrzysztofZurawski\AppData\Roaming\938230184_\06-snap_back.rmed" has been detected as NOT EXISTING : impossible to read anything !

ERROR: In vtkExecutive.cxx, line 741 vtkPVCompositeDataPipeline (00000201601360B0): Algorithm vtkMEDReader (0000020160133950) returned failure for request: vtkInformation (000002016CBF9580) Debug: Off Modified Time: 1694945 Reference Count: 1 Registered Events: (none) Request: REQUEST_INFORMATION FORWARD_DIRECTION: 0 ALGORITHM_AFTER_FORWARD: 1

ERROR: In vtkStreamingDemandDrivenPipeline.cxx, line 819 vtkPVCompositeDataPipeline (00000201601360B0): No update piece number has been set in the information for output port 0 on algorithm vtkMEDReader (0000020160133950).

Exception has been thrown in vtkMEDReader::RequestInformation : File : "C:\Users\KrzysztofZurawski\AppData\Roaming\938230184_\06-snap_back.rmed" has been detected as NOT EXISTING : impossible to read anything !

ERROR: In vtkExecutive.cxx, line 741 vtkPVCompositeDataPipeline (0000020160132AB0): Algorithm vtkFileSeriesGroupReader (00000201600C6110) returned failure for request: vtkInformation (000002016CBFB420) Debug: Off Modified Time: 1707660 Reference Count: 1 Registered Events: (none) Request: REQUEST_INFORMATION FORWARD_DIRECTION: 0 ALGORITHM_AFTER_FORWARD: 1

Exception has been thrown in vtkMEDReader::RequestInformation : File : "C:\Users\KrzysztofZurawski\AppData\Roaming\938230184_\06-snap_back.rmed" has been detected as NOT EXISTING : impossible to read anything !

ERROR: In vtkExecutive.cxx, line 741 vtkPVCompositeDataPipeline (0000020160135F90): Algorithm vtkMEDReader (00000201601350F0) returned failure for request: vtkInformation (000002016CBF7280) Debug: Off Modified Time: 1707691 Reference Count: 1 Registered Events: (none) Request: REQUEST_INFORMATION FORWARD_DIRECTION: 0 ALGORITHM_AFTER_FORWARD: 1

Exception has been thrown in vtkMEDReader::RequestInformation : File : "C:\Users\KrzysztofZurawski\AppData\Roaming\938230184_\06-snap_back.rmed" has been detected as NOT EXISTING : impossible to read anything !

ERROR: In vtkExecutive.cxx, line 741 vtkPVCompositeDataPipeline (0000020160135F90): Algorithm vtkMEDReader (00000201601350F0) returned failure for request: vtkInformation (000002016CBF7280) Debug: Off Modified Time: 1707691 Reference Count: 1 Registered Events: (none) Request: REQUEST_INFORMATION FORWARD_DIRECTION: 0 ALGORITHM_AFTER_FORWARD: 1

ERROR: In vtkStreamingDemandDrivenPipeline.cxx, line 819 vtkPVCompositeDataPipeline (0000020160135F90): No update piece number has been set in the information for output port 0 on algorithm vtkMEDReader (00000201601350F0).
``
This is an ongoing problem since Salome 9.10. Could someone address this? Just want to have possibility to continue work next day from the point where I left it.
Thanks.

[nitawa]

Can you share this file:
C:\Users\KrzysztofZurawski\AppData\Roaming\938230184_\06-snap_back.rmed
as well as the study (hdf file) if any, such that we can check whether this can be reproduced on our side and escalate it?
Thanks

[krzysiek-git]

Hi.
The problem is not study related. It exists at least since v9.10 Windows Native. This was the first version I started using on Windows. Let's reproduce the problem, I'm opening the original study I shared info about:

As I showed you, the study points to some location in Roaming folder. The folder doesn't exist.

So let's point it to the existing results in the correct location:

After loading the results everything is ok again: Let's save the study again: Let's fully close Salome 9.14, run it and reopen the study:

Let's activate paravis:

After activating paravis, the study points to some temporary location instead of the original location. Let's reload the data:

After reloading the data viewport is empty. I tried to open the temporary folder in Roaming... It seems it doesn't exist, as previously.

My guess is that Paravis is using some temporary folder created in Roaming folder and deletes it after closing Salome. Then after reopening the study file for some odd reason Paravis is pointing to that deleted temporary folder while looking for the results. The problem is somewhere while saving the study file. Link to study location is being updated then. This happens to all the studies I'm working currently with. After reopening study files you have to recreate pipeline. Some workaround is saving pvsm file, but this works on random occasion. Paravis sometimes points to good location, but rarely. It points to the temporary location most of the times and I haven't found any pattern to that.

As I said the temporary location doesn't exist after closing Salome. When I hit save this folder containing _smesh.hdf file has been created:

But it stays after closing Salome and it seems it has nothing to do with the results file location. The new folder name is different:

I attached zip containig study etc, but this is not study related. I can recreate the same issue building study from scratch, running simulation and post processing. After saving the study, closing and reopening the result will be the same. Please help.

06-snap_through-arc_length.zip

[krzysiek-git]

I just installed Salome 9.14 on some 10 year old laptop with Windows 10. I opened one of the studies and result is the same. Paravis is pointing to some temporary folder in C:/Users/KrzysztofZurawski/AppData/Roaming/

[krzysiek-git]

Has anyone tested it?

[krzysiek-git]

It's been 2 months since I posted and no one responded. Are you going to work on the issue, or is it a low priority one for the future?