Bug #9392
closedarvbox: Pipeline fails with pip/python errors
Description
I have arvados in a box running and when trying to run a really simple pipeline (which just echo's hello), the pipeline fails and I get the following messages in the log:
09/06/2016 16:27:13 localhost 1 task-print 0 Command /tmp/crunch-job-work/.arvados.venv/bin/python2.7 -c "import setuptools, tokenize;__file__='/tmp/pip-build-y1Mkh1/pycurl/setup.py';exec(compile(getattr(tokenize, 'open', open)(file).read().replace('\r\n', '\n'), file, 'exec'))" install --record /tmp/pip-Q4XQwT-record/install-record.txt --single-version-externally-managed --compile --install-headers /tmp/crunch-job-work/.arvados.venv/include/site/python2.7 failed with error code 1 in /tmp/pip-build-y1Mkh1/pycurl 09/06/2016 16:27:13 localhost 1 task-print 0 Storing debug log for failure in /tmp/crunch-job-task-work/localhost.1/.pip/pip.log 09/06/2016 16:27:13 localhost 1 task-print 0 cat: /tmp/crunch1/src.log: No such file or directory 09/06/2016 16:27:13 localhost 1 task-print 0 /tmp/crunch-job-work/.arvados.venv/bin/pip --quiet install -I /tmp/crunch1/opt/python failed (): exit 1 signal 0 at - line 239. 09/06/2016 16:27:14 localhost 1 task-dispatch 0 child 15652 on localhost.1 exit 111 success=false 09/06/2016 16:27:14 localhost 1 task-dispatch 0 failure (#3, temporary) after 26 seconds 09/06/2016 16:27:14 localhost 1 task-dispatch 0 task output (0 bytes): 09/06/2016 16:27:14 crunch status: 0 done, 0 running, 1 todo 09/06/2016 16:27:14 crunch wait for last 0 children to finish 09/06/2016 16:27:14 crunch Freeze not implemented 09/06/2016 16:27:15 crunch collate 09/06/2016 16:27:15 crunch collated output manifest text to send to API server is 0 bytes with access tokens 09/06/2016 16:27:17 crunch job output d41d8cd98f00b204e9800998ecf8427e+0 09/06/2016 16:27:18 crunch finish
the pipeline looks like this: { "href":"/pipeline_templates/2m5b1-p5p6p-weimjgvp4hs1cnu", "kind":"arvados#pipelineTemplate", "etag":"qg8i3ofech9hnayal65u9c9u", "uuid":"2m5b1-p5p6p-weimjgvp4hs1cnu", "owner_uuid":"2m5b1-tpzed-hs81igix15z88ir", "created_at":"2016-06-09T14:59:55.320505000Z", "modified_by_client_uuid":"2m5b1-ozdt8-fs12c73l29f9auk", "modified_by_user_uuid":"2m5b1-tpzed-hs81igix15z88ir", "modified_at":"2016-06-09T15:25:32.051258000Z", "name":"1st coupel of pipleine steps", "components":{ "bwa-aln-r1-l1":{ "script":"run-command", "script_version":"master", "repository":"arvados", "script_parameters":{ "command":[ "echo", "hello" ], "task.stdout":"test" }, "runtime_constraints":{ "docker_image":"sdgs/pipelinev1", "arvados_sdk_version":"master" } } }, "description":null }
Files
Updated by Matthew Parker almost 10 years ago
- File arvados_issue.txt arvados_issue.txt added
Updated by Matthew Parker almost 10 years ago
I have tried:
Restarting
Rebuilding
Using a vanilla arvados/jobs docker image
I am running:
Ubuntu 14.04 (4.2.0-27-generic)
Maybe I just have to re-install completely?
Updated by Peter Amstutz almost 10 years ago
Hi Matt,
What's happening here is that it is bootstrapping the container by installing the latest Arvados SDK, but failing on the "pycurl" package. However the installation is an optional behavior which is only enabled when "arvados_sdk_version":"..." is provided, so you can try removing that entry from runtime_constraints?