Any temporary output files?

[skill...@gmail.com]

I'm wondering about the run time and if there are log files I can be checking. I started a SMuFin run about 45 hours ago for a matched tumor/normal 30X sample on 15 nodes, 6 cores each, 12 cpus (the cluster is shared so I can't take all cores). Each node has 28GB RAM.

I can see that the process is taking up all the cpu resources, and only a small amount of RAM so that seems to be ok.

The example ran fine.

The log files on the cluster only show this:

"Checking input data files...
BWT index found for /work/projects/synthetic-cancer/public/hg19/index/all.fa"

and poking around in the /tmp directory on the node I can login to doesn't turn up any temporary files or logs that I can see. Are there any, and if so what is the default directory these go to?

[jboy...@gmail.com]

I am having a similar problem (and cannot find error logs)- can you let me know if you have a solution?

thanks

[montse]

Hello,

With a 30X sample, using 16 nodes of 16 cores, the analysis should take arround 8h. About the log file, you can see how is SMuFin going on using a redirection of the error on the cluster.

Regards,
Montse

[skill...@gmail.com]

Yea, I know about redirection most schedulers do that anyhow but the output is not informative. I was wondering if there were interim output files so that we could see how far it had gotten in the analysis process. That would also allow us to use the longer running, but interruptible queue since we are on a shared cluster.

[montse]

SMuFin, at the moment, only provide this information using a redirection of the error on the cluster. We will take into account in next versions.

Thank you,
Montse