CuSe sample input crashes current NOMAD 1.2.2 #95
Assignees
Comments
|
Totally forgot to mention: Docker is docker 25.0.3 on Rocky 9.3: |
|
Hi @behnle, thanks for reporting. Can you confirm you are talking about the |
|
I will take a closer look but I suspect the parsing timed out. However, the error seems to be inconsistent with a timed-out entry. It could also be that the archive size is larger than permitted causing trouble with the archive reader. |
|
Exactly, that's the one. Let me know if you need additional informations for tracking down the problem. |
There are only two parser warnings but no errors. |
|
You can modify the settings by specifying them in the nomad.yaml file. You can have a look at the docs here. For a complete list of config keys, I suggest you look at the code under nomad/config/models.py . For example you can adjust services.api_timeout or celery.timeout |
|
I had already skimmed the list of config options and had set Did not change any celery settings, though. |
|
can you please send me the image path. i have troubles finding it. |
|
Might be related to this: https://gitlab.mpcdf.mpg.de/nomad-lab/nomad-FAIR/-/issues/1861 |
ah yes I have completely forgotten about this. thanks Lauri |
|
@ladinesa What do You mean by "image path"? The docker image? @lauri-codes Yes, might be related to my issue. Would it help to pull a new docker image if available? |
|
@behnle: Sorry I missed your comment. You can try updating the |
|
@lauri-codes Thanks for the heads-up. I recently pulled the "latest" image: With this release, every other attempt on the original sample data succeeds, but some reprocessing runs fail with and the docker-compose log contains error messages like this one In the journal of the server, i found the following potentially related error message: (uid 1000 is the nomad user) I have no clue what is going on. The server has 16 GiB RAM so IMHO an OOM event is rather unlikely (but not impossible). This affects the sample files Edit: NOMAD version is now 1.2.2.dev357+g15b7cd2e1 |
|
@behnle: I will try to reproduce the problem and see why the parser is struggling with this example. |
|
I can confirm that at least one particular main file seems to use a very large amount of RAM, ultimately causing the process to be killed. Here is the zip: int_hse.zip file, the problematic file is We need to check what is causing the memory usage to blow up in the FHI-aims parser for this file. In general some calculations are very big and will need a lot of RAM to be processed, but this does not look like one to me. @ndaelman-hu, @JosePizarro3 : Could you investigate this a bit? |
It's likely this basis set tier checker. I'll see to slim it down. |
|
The issue is Am investigating further. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
When trying to visualize the
CuSe FHI-aims GeometryOptimization simulationsample, NOMAD crashes with a python error:In the GUI, this triggers an internal server error:
Unexpected error: "[object Object] (500)". Please try again and let us know, if this error keeps happening.No cell or workflow graph is shown.
NOMAD version is
1.2.2.dev295+g2e611aff1.The text was updated successfully, but these errors were encountered: