Ansible command line got stuck with shell module

[Hoffmann Liu]

Hi,

I am new to Ansible and immediately fall in love with its simplicity and power.  I have tried Cfengine, Puppet, Chef and even Salt before. I went nowhere with any of them. Now with Ansible I can fast and effectively do what I need.

Today I have a strange issue with the shell module : I wanted to grab some data from remote web access logs :

ansible 192.168.1.1:192.168.1.2  -m shell -a 'zcat  /usr/local/apache/logs/access.log.1.gz  | grep -i fbbd6daa045796e06098132d79'

Then the command just hung there. I checked out both 1.1 and 1.2, the python processes came and went pretty soon and scripts have already exit. But the managing Ansible command just still sat there.

If I changed the grep pattern to something like "index.php" and the results would get back. But sometimes some patterns  just didn't get through.

I uses -vvv to investigate and the last two lines were:

<192.168.1.1> EXEC /bin/sh -c 'LANG=C LC_CTYPE=C /usr/bin/python /root/.ansible/tmp/ansible-tmp-1407579630.71-67715905644258/command; rm -rf /root/.ansible/tmp/ansible-tmp-1407579630.71-67715905644258/ >/dev/null 2>&1'

<192.168.1.2> EXEC /bin/sh -c 'LANG=C LC_CTYPE=C /usr/bin/python /root/.ansible/tmp/ansible-tmp-1407579630.73-200216735938517/command; rm -rf /root/.ansible/tmp/ansible-tmp-1407579630.73-200216735938517/ >/dev/null 2>&1'

Now my workaround is 

ansible 192.168.1.1:192.168.1.2  -m shell -a 'zcat /usr/local/apache/logs/access.log.1.gz  | grep -i fbbd6daa045796e06098132d79  > /tmp/result.txt  2>&1 '

But this is just too tiresome.

I have tried various combinations : from CentOS 5 (ActivePython 2.7.8) to CentOS 5 (stock python 2.4.3) , and CentOS 6 (stock python 2.6.6)  to CentOS 6 (stock python 2.6.6). CentOS5 running with stock openssh4.3 and 6 with stock openssh 5.3.

I have checked syslogs on managing and managed hosts and saw no special messages. I don't know if this is a bug. If so, I 'd love to file a bug report. Is there anything I can do to look into this further ? 

Thanks for your attention.

ywliu

[Petr Sukharev]

Hi!

I check this command in my enviroment and everything is ok.

ansible -i hosts/prod-hosts.ini all --limit="app1" -m shell -a 'zcat /var/log/nginx/access.log-20140806.gz | grep "02:44:20"' show string from file

Why you use old version of ansible? I have 1.4.3 in production and 1.5.X in test.

[Michael DeHaan]

"Why you use old version of ansible? I have 1.4.3 in production and 1.5.X in test."

The latest release is 1.7.0, so you are also running a very old version and should update too :)

For efficiency purposes on this list, we can only really only answer questions about the latest release.

--
You received this message because you are subscribed to the Google Groups "Ansible Project" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ansible-proje...@googlegroups.com.
To post to this group, send email to ansible...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/ansible-project/fcda1313-4502-4443-9fb6-d37c4d301559%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Hoffmann Liu]

Wow. I forgot to mention the most important info  : I use the latest 1.7.0 on CentOS5 managing node and devel-branch on CentOS 6 one.

This is a very weird situation.  It doesn't happen all the time. Is there anything I can do myself to debug this one ?

ywliu