Dear all,
We run our Galaxy server in a virtual machine with less than 1GB RAM. When I'm trying to download results via the Download button in the corresponding history entry that are larger than 1GB, the results file is loaded into memory. This leads to the galaxy process being killed, due to memory issues.
Is there a way to omit loading the result into memory?
Regards, Michael
Hi Michael,
are you compressing the results during download?
Cheers, Bjoern
Dear all,
We run our Galaxy server in a virtual machine with less than 1GB RAM. When I'm trying to download results via the Download button in the corresponding history entry that are larger than 1GB, the results file is loaded into memory. This leads to the galaxy process being killed, due to memory issues.
Is there a way to omit loading the result into memory?
Regards, Michael
Hi Bjoern,
No, I'm not compressing the results. However, Dannon pointed out the problem:
My first guess is that it's likely this is due to 'debug = True' being set in your universe_wsgi.ini. If you set this to false and restart, the entire request will no longer be loaded into memory.
-Dannon
That solved my problem.
Thanks, Michael
On 02/20/2014 01:50 PM, Björn Grüning wrote:
Hi Michael,
are you compressing the results during download?
Cheers, Bjoern
Dear all,
We run our Galaxy server in a virtual machine with less than 1GB RAM. When I'm trying to download results via the Download button in the corresponding history entry that are larger than 1GB, the results file is loaded into memory. This leads to the galaxy process being killed, due to memory issues.
Is there a way to omit loading the result into memory?
Regards, Michael
galaxy-dev@lists.galaxyproject.org