[rgbatista]

Dear someone who might be able to help,




I'm encountering some issues with the process of writing adjoint simulation outputs. Strangely, all my optimization simulations stopped on the 26th around 6:00 PM. Upon investigation, I found that the optimization successfully completes the first direct/primal simulation, and SU2 is able to save/write the resulting files. However, when it proceeds to the adjoint (gradient) simulations and attempts to save/write files at the end of a predetermined number of iterations, SU2 stops and gets stuck, failing to progress beyond this save/write phase after writing "restart_adj_cd.dat". I've conducted optimization simulations from the beginning to the end, including tutorial optimizations, and this is an unprecedented issue for me.


I'll be sharing images related to the file-saving process in SU2 for both direct and adjoint simulations, along with a video of the simulations that came to a halt. I'm puzzled about why it's getting stuck at the file-saving phase.


Link:https://unlpt-my.sharepoint.com/:f:/..._Ojig?e=bPDjDB


https://unlpt-my.sharepoint.com/:f:/..._Ojig?e=bPDjDB


Could this be a lack of space on the server I'm using? In the video I'll share, at the 41-second mark, there's an error indicating insufficient space, "No space left on device," in one of the simulations. I'm including an image as well. I'm utilizing two different directories on a server of 3 TB, one with 270 GB free and another with 282 GB free. Are these storage capacities sufficient? I've experimented with different simulation approaches and sought information on CFD online, but I can't discern why it's unable to save the intended files in the adjoint simulations. The next files it is supposed to write are the PARAVIEW files, but even after removing it from OUTPUT_FILES, the issue persists.


Has anyone ever encountered or heard of a similar problem? Or does know a solution to this problem?




If someone could offer insights or assistance, I would be extremely grateful!


Best Regards,
Rafael Batista

[bigfootedrockmidget]

The error message "no space left on device" is ambiguous and misleading, unfortunately. This message can also appear when you try to write a file in a directory that does not exist, or if you create a subdirectory in a nonexisting directory. It can also appear if you do not have write access to a folder.

[rgbatista]

Hello, bigfootedrockmidget,


Thank you for your quick reply.




So what about stopping at the moment of writing the adjoint files? Is it normal?
My advisor also told me that the server got full and then some space was freed. Could that be the reason? Should I uninstall and install SU2 again? Or can be something related to parallel computation?




Best regards,
Rafael Batista

[bigfootedrockmidget]

So how large are these cases that you try to write? And do you have disk quota for the system that you are using? Maybe just check how much space there is left for you to use and compare it to the size of the data file that is being written.

[itech]

Explore our curated selection of products that encompass the latest trends and advancements in the tech industry. From smart devices to futuristic innovations, i-Tech Groups is your go-to destination for staying ahead in the digital realm of electronic gadgets
and wireless speakers
Our team is dedicated to providing unparalleled customer satisfaction, ensuring that your tech journey is seamless and satisfying. As you navigate through our offerings, remember that i-Tech Groups is not just a platform; it's a commitment to delivering excellence in every technological aspect.

[rgbatista]

Originally Posted by bigfootedrockmidget

So how large are these cases that you try to write? And do you have disk quota for the system that you are using? Maybe just check how much space there is left for you to use and compare it to the size of the data file that is being written.

Usually about 15/20/30 GB, but it depends on the simulation. I have around 200 GB in 2 different directories.


I believe the problem has been resolved by changing the SOLUTION_ADJ_FILENAME and RESTART_ADJ_FILENAME, each with different names, whereas they were identical before. It's strange that I never had a problem when both had the same name.


Anyway, I think it's solved, at least for now.
Thank you for your replies and help.


Best regards,
Rafael Batista