Galaxt training session is slow

Galaxy
1

Hi,

I am currently doing a galaxy training session (gvelasco) and the mapping step of the data is very slow, even if I am using only one chromosome.

Could you solve this issue please ?

Thank you,

Guillaume

2

HI,

I noticed that there are only 16 people actually using the service, while 21 were initially planned. I think the issue might come from that, some participants are probably not included in the reservation, which could explain why their jobs are running more slowly.

Thomas

3

Hello,

It turns out there are fewer students than expected. As far as I know, everyone is connected to the reservation space. I’ll make sure to monitor this more closely moving forward.

Thanks again for your help.

Best regards,
Guillaume

4

Hi,

It seems that for Diffbind tools, 2/16 student have the following error message : “this job was resubmitted to queue because it exceeded the amount of allocated memory on its compute ressource”. I have checked their files and it looks OK.

What can be the problem ?

Thanks

5

This message appears because your jobs are running on reserved nodes, and if several users submit their analyses at the same time, the available memory on the node can become saturated.
When that happens, Galaxy automatically resubmits the job to another queue with more memory.