uniqueg
commented
4 years ago Couldn't be reproduced on c03 cluster with the latest version (changes between #143 and #148), so will close. Re-open if issue still persists or re-occurs
Closed lvarin closed 4 years ago
uniqueg
commented
4 years ago Couldn't be reproduced on c03 cluster with the latest version (changes between #143 and #148), so will close. Re-open if issue still persists or re-occurs
uniqueg
commented
4 years ago Still shows up on c03 when workflow, with auth, is submitted via CLA
lvarin
commented
4 years ago I think the problem is that the $HOME directory of the user in Openshift is /. Should be easily solvable (famous last words).
lvarin
commented
4 years ago After the home directory is created (and using no auth), the error seems to be gone:
$ curl -X POST --header 'Content-Type: multipart/form-data' --header 'Accept: application/json' -F workflow_params='{"input":{"class":"File","path":"ftp://ftp-private.ebi.ac.uk/upload/foivos/test.txt"}}' -F workflow_type=CWL -F workflow_type_version=v1.0 -F workflow_url='https://github.com/uniqueg/cwl-example-workflows/blob/master/hashsplitter-workflow.cwl' 'https://cwlwes.csc-wes.rahtiapp.fi/ga4gh/wes/v1/runs'
{
"run_id": "SKP32H"
}Then,it fails somewhere else:
curl -X GET --header 'Accept: application/json' 'https://cwlwes.csc-wes.rahtiapp.fi/ga4gh/wes/v1/runs/SKP32H'
{
"outputs": {},
"request": {
"workflow_params": {
"input": {
"class": "File",
"path": "ftp://ftp-private.ebi.ac.uk/upload/foivos/test.txt"
}
},
"workflow_type": "CWL",
"workflow_type_version": "v1.0",
"workflow_url": "https://github.com/uniqueg/cwl-example-workflows/blob/master/hashsplitter-workflow.cwl"
},
"run_log": {
"command": "cwl-tes --debug --leave-outputs --remote-storage-url ftp://ftp-private.ebi.ac.uk/upload/foivos --tes https://csc-tesk.c03.k8s-popup.csc.fi/ /data/output/SKP32H/workflow_files/repo/hashsplitter-workflow.cwl /data/output/SKP32H/workflow_files/hashsplitter-workflow.yml",
"exception": "OSError(107, 'Transport endpoint is not connected')",
"expires": null,
"max_retries": 0,
"task_finished": "2020-04-30 05:40:13.031092",
"task_received": "2020-04-30 05:40:13.024255",
"task_started": "2020-04-30 05:40:13.025183",
"time_execution": 0.006,
"time_queue": 0.001,
"time_total": 0.007,
"utc_offset": 0
},
"state": "SYSTEM_ERROR",
"task_logs": []
}Thanks @lvarin! I have never seen this error and unfortunately it's not very informative. Is this on Rahti? What do the Flask and Celery container logs say? Let's take it to Slack
lvarin
commented
4 years ago So the image lvarin/cwl-wes:fake-home should have the changes that fix this ticket. I will activate the auth, solve problems is any and then do a PR :)
Describe the bug
We have been trying the deployment of WES as TESK in Openshift here in CSC in Finland. There seems to be a problem while trying to create a lock file (see the logs below).
To Reproduce Steps to reproduce the behavior:
Go to https://wes-csc-wes.rahtiapp.fi/ga4gh/wes/v1/ui/#/ and run a workflow. @shukapoo has the details on which workflow to run. The workflow ends in error.
Expected behavior
The workflow should finish successfully.
Software Add versions of relevant software, e.g.
Additional context
This is the log of the error: