Closed YUZHENG6 closed 3 months ago
This sounds like a system problem independent of the data. It should be reproducible by running AverageImages by itself
Your system kernel might be too old for the docker image. Can you try an alternative install? Eg, these Centos 7 binaries
This sounds like a system problem independent of the data. It should be reproducible by running AverageImages by itself
I agree with that. And I also tried different versions of GCC. But results will show more errors as same as changed GLIBC version.
Your system kernel might be too old for the docker image. Can you try an alternative install? Eg, these Centos 7 binaries
Actually I used HPC of my insitute and that needs time to update whole system. Are there any other possible solutions to this system problem?
Sorry, I didn't mean you should change the system, rather perhaps the ANTs binaries compiled for Centos 7 might work. They should be built with an older GLIBC
I made a build using an older Debian, does this work for you? If so I can share the Dockerfile
docker pull cookpa/ants:buster
I made a build using an older Debian, does this work for you? If so I can share the Dockerfile
docker pull cookpa/ants:buster
Hi, I apologise for the late reply as I used the Ants you provided, submitted a new task and just got it processed! But in the "shapeupdatedtotemplate", it shows the error that " imagelist length is 18, expected 19". I submitted 19 image files. I'm wondering if it could be an error in processing one of the image files, rather than a problem with Ants, and I need to check with the HPC team.
I made a build using an older Debian, does this work for you? If so I can share the Dockerfile
docker pull cookpa/ants:buster
And I read the error log that shows the "slurm_load_jobs error: Invalid job id specified". I'm wondering if this could be the missing image file. But I read log of each jobs and that show "All_Command_lines_OK".
Sounds like one image failed to register. It might be a slurm problem or something to do with that particular image, you have to find the logs and the job script for that image, and go from there.
Anyway, here's a Dockerfile that builds the most recent ANTs on debian:buster
Operating system and version
Rocky Linux 8.7 (Green Obsidian)
CPU architecture
x86_64 (PC, Intel Mac, other Intel/AMD)
ANTs code version
2.5.2.post5-g3fa6cfb
ANTs installation type
ANTsX docker image (from DockerHub user antsx)
Summary of the problem
When running the command antsMultivariateTemplateConstruction2.sh, errors were reported "AverageImage GLIBC not found"
Commands to reproduce the problem.
!/bin/bash
SBATCH --job-name=antsTemplateConstruction
SBATCH --output=templateConstructionout%j.txt
SBATCH --error=templateConstructionerr%j.txt
SBATCH --ntasks=20
SBATCH --time=4-00:00:00
SBATCH --mem=250G
SBATCH --partition=gpu
SBATCH --gres=gpu:4
SBATCH --mail-user=nassare@crick.ac.uk
SBATCH --mail-type=ALL
antsMultivariateTemplateConstruction2.sh -d 3 -b 1 -n 0 -g 0.15 -q 200x200x100x50 -c 5 -u 50:00:00 -v 100G -o virgin20_ *.nii.gz
Output of the command with verbose output.
Data to reproduce the problem
Data temporarily unavailable