TASK: [name | Copying Build to local host] ************************************
warning: md5sum command failed unusually, please report this to the list so it can be fixed
command: [u’(/usr/bin/md5sum /var/lib/jenkins/jobs/name_qa/workspace/rel/name.tar.gz 2>/dev/null)‘, u’(/sbin/md5sum -q /var/lib/jenkins/jobs/name_qa/workspace/rel/name.tar.gz 2>/dev/null)‘, u’(/usr/bin/digest -a md5 /var/lib/jenkins/jobs/name_qa/workspace/rel/name.tar.gz 2>/dev/null)‘, u’(/sbin/md5 -q /var/lib/jenkins/jobs/name_qa/workspace/rel/name.tar.gz 2>/dev/null)‘, u’(/usr/bin/md5 -n /var/lib/jenkins/jobs/name_qa/workspace/rel/name.tar.gz 2>/dev/null)‘, u’(/bin/md5 -q /var/lib/jenkins/jobs/name_qa/workspace/rel/icda.tar.gz 2>/dev/null)‘, u’(/usr/bin/csum -h MD5 /var/lib/jenkins/jobs/name_qa/workspace/rel/name.tar.gz 2>/dev/null)‘, u’(/bin/csum -h MD5 /var/lib/jenkins/jobs/name_qa/workspace/rel/name.tar.gz 2>/dev/null)']
Thanks Brian.
Ansible 1.5 is an out of date version of Ansible. In particular, there are security fixes you want to be using on the 1.6 line, and the current version is 1.6.10.
Please see if you see this same problem in 1.6.10, and please supply version specifics about your RHEL host.
However, I haven’t /ever/ seen an md5 error from a RHEL host, so this may require some further checking.
The md5 validation flag on fetch is different than the general md5 operations ansible uses to decide if certain files need to be transferred and so on.
It may be true there is a error down the path where flat=yes and fail_on_missing are added, so if you can still reproduce this on 1.6.10 or the development branch, please let us know and we can try to reproduce this.
If that situation is in fact true, filing a ticket at github.com/ansible/ansible would be ideal.
Thanks!
It can happen if the host is in FIPS mode. Not sure this is the case or not.
FYI I updated ansible to 1.6.9 and it appears to be working as expected. There must have been a bug in the past that you guys accidentally fixed. Thanks for the quick response and recommendation…
“accidentally fixed”
That didn’t sound nice
Accidental fixes are the best fixes!
I don’t think a fix was accidental in the least but I do not claim encyclopedic knowledge of 11,467 different commits and when they were applied