My RepeatModeler analysis that I’m running on galaxy.eu has been going for 5 days. Is this normal? Or is the number of threads being used insufficient? Username: alpergokce
Welcome @alpergokce
Yes, analysis of repeats can take some time! The common shared configuration is defined here:
XRef → Hands-on: Masking repeats with RepeatMasker / Masking repeats with RepeatMasker / Genome Annotation
And, for a bit more about how the tool hosting and computational environment works at public Galaxy servers, this was covered in a post last week here → Dada2: removeBimeraDenovo job runtime benchmarking - #2 by jennaj.
Your tool would be similar in that runtimes are about the same in Galaxy as anywhere else given the same inputs, parameters, and resources, plus that the EU server will allow very large jobs to process for extended time periods. So, your choice of server was good!
For the 5 days, I’m not sure if you meant the job was queued (gray) + executing (yellow) for the entire time, or if this is just the executing total days, but I would recommend letting this run to completion. A rerun would only start the whole process over again from the start at the back of the queue. Worst case with a failure, you would at least get the full job logs. These might help when deciding what changes to make should you need to rerun.
Then, checking for how many current jobs for this tool are running:
One of these is probably yours! 
Hope this helps! 