Hi Zheng,
That should not affect memory usage: it probably has to do with
differently assigned nodes on the cluster, which may be shared with
different type of users at that point in time.
S
> Hi Sjors,
> Here I update my running information on the problem data-set. When I
> re-run the problem data-set by setting "Ref.map is on absolute greyscale"
> to No, though the Ref.map were reconstructed by relion, I got through
> finally.
> Thank you for your patient help.
> Regards,
> z.
>
> Zheng Liu Ph.D.
> Postdoc, The Frank Lab
> Department of Biochemistry & Molecular Biophysics
> Columbia University
> http://franklab.cpmc.columbia.edu/franklab/
>
> Zheng Liu Ph.D.
> Postdoc, The Frank Lab
> Department of Biochemistry & Molecular Biophysics
> Columbia University
> http://franklab.cpmc.columbia.edu/franklab/
>
>
> On Mon, Dec 16, 2013 at 3:23 AM, Sjors Scheres
> <[log in to unmask]>wrote:
>
>> Hi Zheng,
>> 32Gb of RAM should do for 384x384, assuming you are only running 1 MPI
>> (+8
>> threads) on each node. Did you actually check this is what is happening
>> by
>> logging onto all nodes and checking how many processors there are?
>> Otherwise, it may be some communication error.
>> HTH, S
>>
>>
>> > Hi Sjors,
>> > Thank you for your reply.
>> > My un-binned image size is 384*384. The RAM of one node is 32G. One
>> > node
>> > have 8 cores.
>> > Zheng
>> >
>> >
>> > On Fri, Dec 13, 2013 at 2:49 AM, Sjors Scheres
>> > <[log in to unmask]>wrote:
>> >
>> >> Probably you've run out of RAM. How much memory do you have? How
>> large
>> >> are
>> >> your images?
>> >> We've seen for example for images of 450x450 pixels that it may take
>> up
>> >> to 15Gb during the iterations, and up to 30Gb of RAM during the last
>> >> iteration….
>> >> S
>> >>
>> >> > Hi Sjors and everyone,
>> >> > When I run auto-refine for my un-binned data, I get following
>> error
>> >> > message for many times so that I can not succeed. But the
>> >> two-time-binned
>> >> > data from the same dataset could run smoothly. Do you ever had this
>> >> issue?
>> >> > And what do you think is the reason of this problem? The version I
>> am
>> >> > using is 1.2.
>> >> >
>> >> > exp_thisparticle_sumweight= 0
>> >> > exp_part_id= 35024exp_iimage=7
>> >> > group_id= 0 mymodel.scale_correction[group_id]= 1
>> >> > exp_ipass= 0
>> >> > sampling.NrDirections(0, true)= 192 sampling.NrDirections(0,
>> false)=
>> >> 192
>> >> > sampling.NrPsiSamplings(0, true)= 24 sampling.NrPsiSamplings(0,
>> >> false)=
>> >> > 24
>> >> > mymodel.sigma2_noise[exp_ipart]= NULL Array
>> >> >
>> >> > [node65:05309] *** Process received signal ***
>> >> > [node65:05309] Signal: Segmentation fault (11)
>> >> > [node65:05309] Signal code: Address not mapped (1)
>> >> > [node65:05309] Failing at address: 0x170
>> >> > [node65:05309] [ 0] /lib64/libpthread.so.0 [0x7ffa2cefdb30]
>> >> >
>> >> >
>> >> > Thank you in advance.
>> >> >
>> >> > Zheng
>> >> >
>> >> >
>> >> > Zheng Liu Ph.D.
>> >> > Postdoc, The Frank Lab
>> >> > Department of Biochemistry & Molecular Biophysics
>> >> > Columbia University
>> >> > http://franklab.cpmc.columbia.edu/franklab/
>> >> >
>> >>
>> >>
>> >> --
>> >> Sjors Scheres
>> >> MRC Laboratory of Molecular Biology
>> >> Francis Crick Avenue, Cambridge Biomedical Campus
>> >> Cambridge CB2 0QH, U.K.
>> >> tel: +44 (0)1223 267061
>> >> http://www2.mrc-lmb.cam.ac.uk/groups/scheres
>> >>
>> >>
>> >
>> >
>> > --
>> > Zheng Liu Ph.D.
>> > Postdoc, The Frank Lab
>> > Department of Biochemistry & Molecular Biophysics
>> > Columbia University
>> > http://franklab.cpmc.columbia.edu/franklab/
>> >
>>
>>
>> --
>> Sjors Scheres
>> MRC Laboratory of Molecular Biology
>> Francis Crick Avenue, Cambridge Biomedical Campus
>> Cambridge CB2 0QH, U.K.
>> tel: +44 (0)1223 267061
>> http://www2.mrc-lmb.cam.ac.uk/groups/scheres
>>
>>
>
--
Sjors Scheres
MRC Laboratory of Molecular Biology
Francis Crick Avenue, Cambridge Biomedical Campus
Cambridge CB2 0QH, U.K.
tel: +44 (0)1223 267061
http://www2.mrc-lmb.cam.ac.uk/groups/scheres
|