[openstack-ansible] strange execution delays
On Mon, Dec 30, 2019 at 12:00 AM Joe Topjian <joe at topjian.net> wrote:
> I don't think this is an OSA issue per-se, but it's happening on infrastructure deployed with OSA.
> Using both the Rocky and Stein releases of OSA on Ubuntu 18.04, we're running into a strange a strange issue where the run times of Ansible will eventually take longer and longer. Digging into the issue more, we find that it's when Ansible executes a command within a container and the call to "su" triggers the delay. Over time, the delay increases more and more. The same behavior is seen if we SSH into containers.
> The only workaround we've been able to identify that works is to stop and start the containers.
> We have not been able to identify a root cause of this, nor any consistency to what containers this will happen on.
> Has anyone run into this?
No, this is pretty strange.
Do you have any PAM modules that might be hitting some sorts of
external API for auditing purposes that may be throttling you?
How is systemd-logind feeling? Anything odd in your system logs?