Open cymon opened 1 month ago
I am not sure I am following; is this when you are running the get_singularity_images.sh script?
Where is this docker://docker:/
pattern you mention?
I guess you can try to get the images in any possible way and then use them from cahce.
I am not sure I am following; is this when you are running the get_singularity_images.sh script?
No, this is when running metaGOflow itself.
Where is this
docker://docker:/
pattern you mention?
This is in the python traceback or the error:
FATAL: While making image from oci registry: error fetching image to cache: failed to get checksum for docker://docker://cymon/eggnog-2.1.12:0.2: unable to parse image name docker://docker://cymon/eggnog-
8018 ^[[1;30mERROR^[[0m ^[[31mGot workflow error: Singularity is not available for this tool, try --no-container to disable Singularity, or install a user space Docker replacement like uDocker with --user-space-
8019 Traceback (most recent call last):
8020 File "/share/apps/lib/python3.7/site-packages/cwltool/job.py", line 809, in run
8021 runtimeContext.tmp_outdir_prefix,
8022 File "/share/apps/lib/python3.7/site-packages/cwltool/singularity.py", line 307, in get_from_requirements
8023 if not self.get_image(cast(Dict[str, str], r), pull_image, force_pull):
8024 File "/share/apps/lib/python3.7/site-packages/cwltool/singularity.py", line 249, in get_image
8025 check_call(cmd, stdout=sys.stderr) # nosec
8026 File "/share/apps/lib/python3.7/subprocess.py", line 347, in check_call
8027 raise CalledProcessError(retcode, cmd)
8028 subprocess.CalledProcessError: Command '['singularity', 'pull', '--force', '--name', '/home/cymon/src/metaGOflow-cymon.git/sif_images/cymon_eggnog-2.1.12:0.2.sif', 'docker://docker://cymon/eggnog-2.1.12:0.2
8029
I guess you can try to get the images in any possible way and then use them from cahce. I have the images in the sif_images folder and they work when running the exact same code on different nodes; but just these 3 node give this error - and I can't see why...
I have not seen this error before but based on the error message, is there any chance singularity is not installed correctly in those nodes?
I have some serious weirdness that I thought I should document.
I have a small HPC; head node, 6 compute nodes, compute nodes mount /share/apps /home from head node. I have metaGOflow installed in /share/apps - so all nodes see the same code. All nodes run an up to date AlmaLinux release 8.10. All nodes have
apptainer.x86_64 1.3.2-1.el8
from EPEL repo installed andpodman-docker.noarch
plus dependencies. No other conflicting container software is installed.The problem: some nodes fail (not all just 3 out of 6) to run metaGOflow giving the following error:
The call being attempted is:
It looks like the
docker://docker://
string is being mis-formed, and should only have onedocker://
This happens for other images as well, but only on certain nodes:
Google does not help at all. No idea what is going on...
I'm assuming this is not a problem with the workflow but the configuration of some of my nodes... but thought it put it here for posterity anyway...