[chef] Re: Re: Re: Re: Chef client Zombie

[Astlock < >]

I've tested on a different sizes of instances and in one of "stop/start" the issue disappeared. It is really strange, do not want to get it into production.

On Thu, May 23, 2013 at 8:12 PM, Patrick Gardella < " target="_blank"> > wrote:

We've had this problem for some time. I thought it was fixed under 11.4, but it happened again this week. I think Elvin's comments are spot on in my scenario. One of our DNS servers was not responding to queries externally, and the other one crashed.  So it took all my chef-client managed nodes out.  Now that I've fixed the DNS servers, we've not had a problem.  And more to the point, I would have said the DNS wasn't to blame before this week either.

Patrick

On Thu, May 23, 2013 at 1:00 PM, Astlock < " target="_blank"> > wrote:

Thank you Elvin. The problem is not in the DNS it is working correctly. Thanks for the info.

 

On Thu, May 23, 2013 at 7:46 PM, Elvin Abordo < " target="_blank"> > wrote:

i've seen this before in my setup. turned out to be an inaccessible DNS server in /etc/resolv.conf. Although my environment wasn't in EC2 and was debian6. so YMMV

On Thu, May 23, 2013 at 12:45 PM, < " target="_blank"> > wrote:

I have encountered a problem with Chef: 11.4.4 and ec2 ubuntu-12.04. In one of
a new node after each run of chef-client added 7 zombie processes.

--
Elvin Abordo
Mobile: (845) 475-8744