Webserver stall

Good morning all -

My colleague and I started a few runs last Friday (the 15th), they seem to have stalled (they have not moved in “% done” in a few days). Is there the ability to download the files that have been created so far so we can what the problem is?



We don’t see any active run from Friday 15th on our servers…

Dr. Bonvin -

Not sure what is going on then. Here are the job names of the runs:


One is stalled at 100% on it1, 0% on itw
The other is stalled at 99.5% on it1, 0% on itw



Found them!

It looks like you have hit a feature. The current code only supports 5 sets of RDCs…

I tweaked your runs to make them continue. But if you submit a similar run it will hang again until we fix this issue.

Dr. Bonvin -

Thanks for this! One more thing though, the “ALLRDC_SPLIT_AMBIG1_VEAN_SANI” run is only 23MB when downloaded and there is no “structures” directory. Is it salvageable?


Something must have gone wrong in the post processing… try later to see if the archive has changed- if not then I am afraid it is lost

We should soon have a fix to support up to 10 RDC sets

Hi Ben,

The server has been updated to support 10 RDC restraint sets.

Can you please test if all goes fine with your system?

Dr. Bonvin -

It does not appear to be fixed. It has been stalled at it1 at 99.5% for a while.

Here is the name of the run: all_rdc_ambig_1_vean_sani



That’s a different issue - you have models failing …

Actually the problem comes from the tensors which are not yet used in the first stages and float away, causing numerical problems in your case.
Another feature that should be fixable.

Hi Ben

I successfully rerun your run (without errors) - you seem to have had bad luck…

Anyway we now implemented a fix to make sure the tensor moving away is not leading to numerical issues.

Give it another try!