yum update hangs with name argument '*'
#51284
Comments
|
Files identified in the description: If these files are inaccurate, please update the |
ansibot
added
affects_2.7
|
Here is the same with I also tried logging in manually via ssh and executing |
|
The command that hangs is executed in https://github.com/ansible/ansible/blob/v2.7.6/lib/ansible/modules/packaging/os/yum.py#L641 The following reproduces the freezing effect: Edit: The command above actually finished after quite a while (1-2 minutes). It would still be nice if the yum module handled such "erroneous" configurations so the user is not confronted with a frozen process. |
|
The following line seems to be causing the issue: |
|
@Timoses I don't have a system available to me that I'm able to reproduce this with, is there a certain set of public yum repositories you're using that adds up to that large of a content set for needs_info |
ansibot
added
needs_info
|
@maxamillion The host is served by a rpm repository(s) managed by spacewalk. That means there are several "channels". I'm afraid I don't have any more information, at least not as insight I can give you as I myself am not well versed with yum and rpm. |
ansibot
removed
the
needs_info
|
@Testicule is sounds like you are just invoking the module without specifying the args json file that contain the arguments. If no argument is specified it will try and read from stdin. You can read all about how to debug/invoke modules outside of Ansible at https://docs.ansible.com/ansible/latest/dev_guide/debugging.html. |
|
@Testicule |
|
@Timoses no worries, I'll try and see if we can sort it out. I'm curious how many repos you have listed in needs_info |
ansibot
added
the
needs_info
|
@maxamillion Sorry for the delay. (blanked out the channel names with |
ansibot
removed
the
needs_info
|
@Timoses no worries, thank you! Out of curiosity, if you just run a Thanks again. needs_info |
ansibot
added
the
needs_info
|
@maxamillion
Running Upon running With Nothing happens (at least not visible on the ansible controller) after the second /etc/yum.conf (in case it helps?) (Is there any way to print the commands that an ansible module executes? This would be useful for reproducing issues.. It's a pain debugging the python modules live : /.) |
ansibot
removed
the
needs_info
|
@Timoses this is great, thank you! I'll dig in and see if I can sort out the root cause. |
|
@maxamillion : Any updates on this? The documentation of the yum module is still striking me as a bit odd. Under the field
It does not say that state may be anything else. Does this mean Please also see the discussion here on the application of a use case with The problem still persists, also when explicitly setting |
|
@Timoses unfortunately not, I've been under water with other priorities ... I haven't forgotten about it, but I've just not had a chance to circle back to it. Apologies. |
|
was this ever addressed. Having the exact same issue all of a sudden |
|
I actually think this issue was related to #63713 ... I suspect what's happening is that it's attempting to re-instantiate the YumBase (and subsequent sack load, which is an extremely expensive operation) too many times on a large package count and timing out. That fix has been merged and there are backport pull requests open so I would hope to see this fixed in the next releases of both Ansible 2.8 and Ansible 2.9. Please let me know if you have a chance to test and verify. |
|
I am having the same issue in Ansible v2.9.2 using:
Any update on a resolution to this? |
|
Files identified in the description: If these files are incorrect, please update the |
|
@ses-team I need more information about the system so that I can reproduce and diagnose.
needs_info |
ansibot
added
the
needs_info
|
@Timoses This issue is waiting for your response. Please respond or the issue will be closed. |
|
@Timoses You have not responded to information requests in this issue so we will assume it no longer affects you. If you are still interested in this, please create a new issue with the requested information. |
and others
SUMMARY
yummodule with argumentname: '*'hangs indefinitely.ISSUE TYPE
COMPONENT NAME
yum
ANSIBLE VERSION
CONFIGURATION
OS / ENVIRONMENT
Operating System: CentOS Linux 7 (Core)
ethtool -i eth0
driver: virtio_net
version: 1.0.0
STEPS TO REPRODUCE
EXPECTED RESULTS
Ansible finishes and succeeds or reports failure message.
ACTUAL RESULTS
Running with
ansible-playbook -vvv --become --become-user=root -i inventory/hosts.ini playbook.yml:I can simply log onto the machine and do
sudo yum -y updateand it works.The yum repository used is a privately hosted one.
The text was updated successfully, but these errors were encountered: