Mia,

also i noticed this at the top of the log:

total used free Mem: 37010736 19490372 17520364 -/+ buffers/cache: 10784236 26226500 Swap: 8393952 150528 8243424

it shows that just 1.7GB is available for the recon to run, which is below what we recommend. recon-all usually needs at least 2GB.

after inspecting for defects like doug suggests, you could try stopping other apps that are consuming mem (or just reboot which will ensure maximal free memory) and then run recon-all -make all

Nick

On Wed, 2013-01-23 at 13:23 -0500, Douglas N Greve wrote:

Hi Mia, try looking at the lh.orig.nofix surface. There may be a huge defect (like the entire cerebellum is attached). You can run

tkmedit MG58_SurferOutput_postop orig.mgz lh.orig.nofix

Note that you will get a few error messages when loading the orig.nofix surface. These can be ignored

doug

On 01/23/2013 12:19 PM, Borzello, Mia wrote:

...

Hey Freesurfers,

I received a "Cannot allocate memory" error while running a recon overnight. I think there should be enough space on the machine to store the data so I'm suspicious that it's a different memory problem. I attached the recon-all.log file.

Thanks for your help! m

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Mia,

this time it looks like there was plenty of memory, but it exited on a defect. make sure the wm.mgz looks good (doesnt have stuff outside the white matter). see the 'topological defects' section of our tutorial for info on what to look for and how to fix it.

Nick

Hi freesurfers,

So i reran recon-all -make all overnight, and it ended with errors still. It looks like the same error (I attached the log). There was another process running while I ran the recon the first time, but I made sure that wasn't the case this time around. Is the solution running the recon in a whole new place with more space?

Thanks so much, m ________________________________________ From: freesurfer-bounces@nmr.mgh.harvard.edu [freesurfer-bounces@nmr.mgh.harvard.edu] on behalf of Nick Schmansky [nicks@nmr.mgh.harvard.edu] Sent: Wednesday, January 23, 2013 1:45 PM To: Douglas N Greve Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Memory problem

Mia,

also i noticed this at the top of the log:

         total       used       free

Mem: 37010736 19490372 17520364 -/+ buffers/cache: 10784236 26226500 Swap: 8393952 150528 8243424

it shows that just 1.7GB is available for the recon to run, which is below what we recommend. recon-all usually needs at least 2GB.

after inspecting for defects like doug suggests, you could try stopping other apps that are consuming mem (or just reboot which will ensure maximal free memory) and then run recon-all -make all

Nick

On Wed, 2013-01-23 at 13:23 -0500, Douglas N Greve wrote:

...

Hi Mia, try looking at the lh.orig.nofix surface. There may be a huge defect (like the entire cerebellum is attached). You can run

tkmedit MG58_SurferOutput_postop orig.mgz lh.orig.nofix

Note that you will get a few error messages when loading the orig.nofix surface. These can be ignored

doug

On 01/23/2013 12:19 PM, Borzello, Mia wrote:

...

Hey Freesurfers,

I received a "Cannot allocate memory" error while running a recon

overnight. I think there should be enough space on the machine to store the data so I'm suspicious that it's a different memory problem.

...

I attached the recon-all.log file.

Thanks for your help! m

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer