A few days ago, I submitted a job with the tool ‘NCBI FCS GX’ but it hasn’t started yet. It should be a simple task. Is the public EU Galaxy server really that backed up?
Is your job still queued? The queue is sorting jobs between available cluster nodes based on technical specifications appropriate for the estimated needs. Smaller nodes process jobs faster and they might be busy due to being in very high demand most of the time. Larger nodes process longer running jobs so the turn over isn’t as fast, and are where the day-to-day queue time variation tends to show up.
For context, these are the is the baseline node sorting specifications for NCBI FCS GX. It seems it isn’t routed to the smaller nodes. It is going to a node with more cores and runtime memory. A job with a smaller input and a larger input will queue with the same timing. Why? The size of the reference index involved is very large and is the primary reason for the scaled up resources.
Next, we can check the server statistics → How to see the UseGalaxy.eu job queue statistics
There are about 15 jobs queued for this tool. Your job is likely in this count!
Please let us know how this turns out! If you need any help confirming that the inputs are ready to use or that the job is actually queued and not paused, you are welcome to share the history with the job or some screenshots of the same details. → FAQ: Troubleshooting errors
Thanks and we can follow up more! 
Then, just for notice here, there is a planned administrative maintenance next week. Everyone should get this pop-up.
Thanks, @jennaj. Yes, the job is still queued. I started another job with the same tool at about the same time, and that one started after several hours and finished quickly. So, I thought maybe the other job would run soon as well, but it hasn’t yet after another couple of days.
Ok, thanks for the extra context. I am wondering if your original job was missed or unscheduled for some reason.
I would suggest trying to rerun that same job all over again (with the rerun button) but still leave the original just in case. Then, if/when the rerun completes, then you can go ahead and delete the original.
Reruns like this are big picture a “poor strategy” since it creates more stress on the server with duplicated runs clogging up queues, but might be Ok in a few corner cases! And with this particular server pending some infrastructure updates, I would expect situations like this to become (even more) rare.
Let us know what happens!
It finally finished! The second run had started but I cancelled it. Thanks, @jennaj 