GISTIC2.0 error

23 views
Skip to first unread message

Swetha Ramadesikan

unread,
Jan 31, 2023, 12:13:29 PM1/31/23
to GenePattern Help Forum
Hello,
I have been trying to run the GISTIC2.0 module on the public cloud server. I have previously run this with no issues. Since last night, I see the following error in my error log:
"Essential container in task exited"
"write /var/lib/docker/tmp/GetImageBlob351967463: no space left on device". 

The input files have not significantly changed and I tried to rerun an older input file that had previously successfully generated outputs. Some of my JOB IDs are 490338, 490137, 490145.

Thanks for your time and assistance!
Swetha

Ted Liefeld

unread,
Jan 31, 2023, 1:07:30 PM1/31/23
to GenePattern Help Forum
Swetha,

you can retry your job now and it should be able to go through, the problem was one of our compute nodes ran out of disk bit failed to shutdown.

Ted

Long version of what failed:

We dispatch jobs on AWS and they can share compute nodes.  Another user ran a job that consumed the entire disk (~170GB) but the process did not die, it just sat in a hung state.  The scheduler saw that the CPU on this node was 0% and so it kept dispatching jobs to this compute node, which then failed because there was not enough space to get the docker container or input files.

Swetha Ramadesikan

unread,
Feb 1, 2023, 9:39:29 AM2/1/23
to GenePattern Help Forum
Ted, thank you so much for your quick response!! That fixed it, I am able to successfully complete jobs now!

Swetha

Reply all
Reply to author
Forward
0 new messages