hi jens,
if you search the archives, there is a post that lists the minimal set of files needed for running recon-all.
i used a tool called cde to create it: http://www.pgbovine.net/cde.html
this reduces the archive down to about 650MB comprising binaries and atlases.
this wouldn't be for general purpose freesurfer processing, just for running recon-all.
cheers,
satra
On Wed, Nov 19, 2014 at 8:40 AM, Jens Offenbach wolle5050@gmx.de wrote:
I have corresponding setup modules for auto-deployment and configuration in the cloud, but it still takes some time (even in the local cloud network) to transfer the archive from the fileserver to the virtual machine and especially to untar that archive. A manual untar took nearly 20 minutes, that's too slow for on-demand provisioning. I think the "subjects" folder can be removed. It seems to contain only example mri data, but I want to make sure not to remove anything that is actually required for analysis purposes. I need a minimum archive file which is able to perform the pipeline execution.
Yes, we have a cloud portal, where different consumers can request the freesurfer service after uploading their mri raw files. The virtual machines are created and configured on the fly starting from a raw OS image. When the freesurfer run terminates the virtual machines are destroyed and the whole procedures starts again if someone requests a new freesurfer service (Create and start a virtual machine, transfer freesurfer archive, untar that archive, set environmement variables ...). We only want to parallelize different freesurfer jobs and not the freesurfer pipeline itself. We do not have a medical background.
Regards, Jens
Gesendet: Mittwoch, 19. November 2014 um 13:14 Uhr Von: "Krieger, Donald N." kriegerd@upmc.edu An: "'Freesurfer support list'" freesurfer@nmr.mgh.harvard.edu Betreff: Re: [Freesurfer] Pipeline/Step Progress Monitoring, Packaging and 5.3.0-patch Dear Jens,
I can offer just a bit of information and I also have a question. We are using Fedora 11.0 on an x86_64 box and both the Centos 4 and Centos 6 compiles work. The package appears to be self-contained as I have seen no faults on failures to find shared libraries. We simply untar freesurfer to a convenient location and set the PATH and FREESURFERHOME environment variables accordingly. I see no reason why that shouldn't work in a cloud solution except that I do not know what you have in mind as a use for your solution. Will there be a group of users who are provided access so that they can run freesurfer on an image data set through a portal to your cloud? Are you using the cloud to distribute the calculation to many machines? If so, what is the "granularity" of this, e.g. does one freesurfer run get one machine?
Regards,
Don
Don Krieger, Ph.D. Department of Neurological Surgery University of Pittsburgh (412)648-9654 Office (412)521-4431 Cell/Text
-----Original Message----- From: freesurfer-bounces@nmr.mgh.harvard.edu [mailto:freesurfer- bounces@nmr.mgh.harvard.edu] On Behalf Of Jens Offenbach Sent: Wednesday, November 19, 2014 5:47 AM To: freesurfer@nmr.mgh.harvard.edu Subject: [Freesurfer] Pipeline/Step Progress Monitoring, Packaging and
5.3.0-
patch
We are currently working on a solution that brings FreeSurfer in the
cloud.
Unfortunately, we are facing a lot of problems. Hopefully, someone can
help
me:
I want to run FreeSurfer on Ubuntu 14.04 or 14.10. What package is
required to
download (CentOS 4 x86_64 (64b))? Does it work anyway on the Ubuntu
family?
On the FTP server I have found a patch for 5.3.0, but I am not sure how
to apply
it or if I must apply it? The Readme says to replace “mri_segstats”.
What about
the other files “recon-all, fsfast, tksurferfv, …”?
I want to reduce the archive to its bare minimum, otherwise the automated deployment process takes too long. It should contain only those files
which are
necessary to execute the pipeline without any viewer or examples. Which
files
and folders can be deleted safely?
The most important issue is to monitor the FreeSurfer pipeline
execution. I know
that the scripts folder contains some status files which offer important
data
helping me to monitor the pipeline execution progress of “recon-all”. Unfortunately, that is not enough… If one step takes e.g. approximately
20
hours, I have to say something about the progress within this current
step. Is
there any possibility to monitor the progress of a single step. We need something that allows us to say e.g. that Step “CA Reg” has 34 % of its
work
done so far which means that the overall pipeline progress is e.g. 63 %.
Is it
possible to say that a specific step takes a specific portion of the
overall
pipeline execution time. What influences this portion (format or quality
of input,
platform, …)?
Thank you very much for your help!
Regards Jens
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom
it is
addressed. If you believe this e-mail was sent to you in error and the
contains patient information, please contact the Partners Compliance
HelpLine
at
http://www.partners.org/complianceline%5Bhttp://www.partners.org/compliancel...] . If the e-mail was sent to you in
error but does not contain patient information, please contact the
sender and
properly dispose of the e-mail.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer%5Bhttps://mail....]
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer