Hi,
I tried to launch the decontamination tools NCBI FCS GX and it stay in waiting mode since days. When I launch another job with another software it works fine. I don’t understand what the problem is.
1 Like
Welcome @GuillaumeDore11
Most tools will queue before running but several days does seem odd. Different tools route to different clusters, so some might have appropriate clusters available sooner that others. More details are in the topics here with the queued-gray-datasets tag.
I’ve started up some simple tests to see what happens.
- EU server https://usegalaxy.eu/u/jenj/h/test-ncbi-fcs-gx
- ORG server https://usegalaxy.org/u/jen-galaxyproject/h/imported-from-archive-test-ncbi-fcs-gx
You are also welcome to share back yours. If the timestamps really do show a multi-day wait, we can reach out to the administrators to get feedback about what might be the issue. The server usually matters, and the share link will include it but if you capture screenshots instead please do clarify the server.
Let’s start there, thanks! 
Hello,
Thanks for your reply. Here is a link for a test that I just launch : https://usegalaxy.eu/u/guillaume_dore/h/test-ncbi-gx-fcs
I’m also sending you a screenshot of one job that I launch on Tuesday on EU server. I tried launch jobs already on Monday but I stopped them because they weren’t launching.
Thanks for sharing that back! Very odd. My job at the EU server is still queued, too, and it failed without any details at the ORG server.
I’m getting some feedback right now, more soon!
The tool should be working at UseGalaxy.org now if you want to try it there. How to move data between servers → FAQ: Transfer entire histories from one Galaxy server to another
Why this is stalled for so long at UseGalaxy.eu is not clear and likely something different. Let’s ping one of the local administrators: hi @wm75 would you be able to check for us? Thanks!
1 Like
So the reason why jobs for this tool get queued so long on Galaxy Europe is that we have quite enormous resource requirements (480GB RAM and 64 cores) defined for them at least in screening mode.
@jennaj you could help us by telling us what was configured on usegalaxy.org to enable a successful run.
Thanks again for reporting the issue and for sharing your example.
After reviewing, we confirmed that both servers have the same tool version and resources allocated. The difference is that the EU server’s really large clusters (needed for this tool) happen to be very busy right now. This means that a job like yours that runs on the largest clusters will queue for a much longer time. The size of the inputs do not really matter – the resource use is mostly related to the massive indexes.
What to do
Please leave these jobs queued and avoid re-running.
All new jobs are added back to the end of the queue again. If deleted and rerun often enough, the job may never get a chance to run!
My test job from Wednesday at the EU server is still queued too. I’ll watch to over the weekend and your history too if you leave it shared.
Job execute as resources become available so it is difficult to predict the exact timing. Much depends on the cluster load at any particular time. By chance, the ORG server was “faster” but don’t rely on that always being true.
Tools that utilize the very largest clusters will queue the longest. Why? Because the currently processing jobs tend to take longer to complete which leads to slower turnover for new jobs.
I’ll also leave this quote below as a reference for anyone that happens to read this topic later on.
Thanks again and feel free to ask for an update next week! 
Thanks for your answer. I launched it on http://usegalaxy.org/ and it worked very quickly.
Thanks again !
1 Like
Is there still any chance to resolve this issue on https://usegalaxy.eu/, or every time we need to move data to the usegalaxy.org server?
Are you having trouble with the tool right now? The tool was working at the EU server when I had tested back in April. I just started up another test job in the same history to see what happens, but I am expecting it to also process Ok.
Now, these jobs were sometimes queuing for a longer time period than expected. I can’t recall exactly what we did last April but I’m fairly certain resource changes were applied (@wm75 will see this and can comment more if I misunderstood?).
For job timing now, we can let you know that the winter holiday period is the busiest time of the year at all of the Galaxy servers! Many jobs and tools will queue a bit longer due to the extra traffic as researchers are wrapping up projects or learning new methods.
Remote sites like NCBI are probably busy as well. And, the underlying computational resource sites tend to have their own maintenance cycles over the larger academic breaks or at the end of the year. This should all resolve over the next week or so as everyone is getting back to their usual routines and the school year starts up.
The best advice we have is to get jobs into the queue then to allow them to process! Odd failures may resolve with a rerun, so please try at least one. Persistent errors can be reported and we’ll try to correct any outstanding issues that may come up. If you notice unusual timing, this can also be reported (an example job is usually helpful).
Hope this helps to clarify the usage and please let us know if you notice anything unexpected! 
I submitted job, it run after four days on usegalaxy.eu
However it was working quickly on usegalaxy.org server.
1 Like
Ok, thanks for that feedback! Maybe there is something else we can do. More soon!