[aspect-devel] More vtu problems

Wed Aug 2 17:05:24 PDT 2017

   Lev,

       So I guess there must be something wrong with the Aspect output file.

       Try to check if you can open other pvtu files in Paraview, in order
to rule out a crash of this latter.
        Could you resume the computation in Aspect from the latest, or
second latest checkpoint? Cause the problem might have arisen after those..

cheers,
Felipe

On Wed, Aug 2, 2017 at 1:54 PM, Lev Karatun <lev.karatun at gmail.com> wrote:

> Thanks everyone for the answers!
>
> John, I'm only opening a single file. As a small test, I closed most apps
> so that I only had 3/16 GB used, then opened one timestep file. Memory
> consumption spiked to 10.7GB. Then, when I tried to plot the velocities, it
> crashed. So I guess Timo might be right, paraview may be going OOM.
>
> Felipe, I tried switching to just outline for the actual model, and
> plotting every 10000th velocity vector, it still crashed.
>
> Quite frustrating not to be able to work with a single 300MB file on a 16
> GB laptop.
>
> Best regards,
> Lev Karatun.
>
> 2017-08-02 15:05 GMT-04:00 John Naliboff <jbnaliboff at ucdavis.edu>:
>
>> Hi Lev,
>>
>> When you open the paraview view file are you opening the solution.pvd
>> file (loads all time steps) or the individual time steps (.pvtu files)
>> separately?
>>
>> If you are opening the solution.pvd file, I would be curious to see if
>> the error occurs for the same time steps if you only load the single .pvtu
>> file. This would at least help pin-point if this is a "paraview" problem,
>> rather than something inherent with the files themselves.
>>
>> My first guess would have been that you ran out of memory as well, but
>> that is clearly not the case. I have had paraview crashes on my laptop (16
>> Gb) with large models and large amounts of output data, but I don't
>> remember the exact error messages.
>>
>> Cheers,
>> John
>>
>> *************************************************
>> John Naliboff
>> Assistant Project Scientist, CIG
>> Earth & Planetary Sciences Dept., UC Davis
>>
>> On 08/02/2017 11:21 AM, Lev Karatun wrote:
>>
>> Hi,
>>
>> I ran into more problems trying to open vtu files. This time I used the
>> option "number of grouped files = 1", and the output file opened without
>> errors, but when I try to visualize the velocity using glyphs in Paraview,
>> it crashes. It doesn't crash on a couple of first output files, only on the
>> later ones. I attached the screenshot with error message. It's paraview 4
>> on the screenshot, but I tried the earliest version with little luck.
>> I tried doing the same in Visit, it didn't crash, but produced an error
>> and didn't plot velocities either.
>> I also tried checking the vtu file for NULL characters. Obviously I can't
>> check a 300 mb file with bare eye, but I tried doing
>> *tr < solution-00004.vtu -d '\000' > sol-04*
>> and there was no difference between the original file and output of tr. I
>> tried the same command the files I had problems with earlier, and there was
>> difference, so I assume the command is correct.
>> From the error message, I thought maybe Paraview ran out of memory, but I
>> actually still have about about 4GB free. Also, the problem persists when I
>> open a single 300mb file, which I successfully did many times in the past.
>>
>> If someone has any ideas on how to solve this, I would appreciate it.
>>
>> Best regards,
>> Lev Karatun.
>>
>>
>> _______________________________________________
>> Aspect-devel mailing listAspect-devel at geodynamics.orghttp://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>>
>>
>>
>> _______________________________________________
>> Aspect-devel mailing list
>> Aspect-devel at geodynamics.org
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>>
>
>
> _______________________________________________
> Aspect-devel mailing list
> Aspect-devel at geodynamics.org
> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.geodynamics.org/pipermail/aspect-devel/attachments/20170802/09d758b2/attachment-0001.html>