Hello,
I am running galaxy, using a postgres backend.
The galaxy server systematically gets killed whenever I try uploading a bam file. The samtools process is still running. This is the end of the log:
2010-08-26 18:08:11,371 INFO sqlalchemy.engine.threadlocal.TLEngine.0x...4610 {'param_1': 8} sqlalchemy.engine.threadlocal.TLEngine.0x...4610 INFO 2010-08-26 18:08:11,371 {'param_1': 8} 2010-08-26 18:08:11,378 INFO sqlalchemy.pool.QueuePool.0x...4ed0 Connection <connection object at 0x567bf30; dsn: 'dbname=galaxy host=/var closed: 0> being returned to pool sqlalchemy.pool.QueuePool.0x...4ed0 INFO 2010-08-26 18:08:11,378 Connection <connection object at 0x567bf30; dsn: 'dbname=galaxy host=/var closed: 0> being returned to pool 10.1.20.44 - - [26/Aug/2010:18:08:06 -0700] "POST /root/history_item_updates HTTP/1.1" 200 - "http://galaxy:8080/history" "Mozilla/5.0 (X1 (x86_64); en-US; rv:1.9.2.8) Gecko/20100722 Firefox/3.6.8" Killed
I have the following options set in universe_wsgi.ini:
database_engine_option_server_side_cursors = True database_engine_option_strategy = threadlocal
It seems to be running out of swap. I monitored memory usage with 'top' and galaxy died when available swap hit 0k free.
I am not superuser on the machine where galaxy is running and I have no other server available. Is there something I could do to prevent galaxy from dying?
Thanks,
Isabelle
-- Isabelle Phan, DPhil Seattle Biomedical Research Institute +1(206)256 7113