Fact gathering times out

[E.C. Raymond]

I am seeing a problem with Ansible (1.9.4) and Debian 7 (wheezy) timing out and failing on gathering facts, specifically with autofs/network mounted volumes. Its not limited to a specific host, as it is happening at random times across 200 hosts that have autofs setup. 

GATHERING FACTS ***************************************************************

failed: [localhost] => {"cmd": "/bin/lsblk -ln --output UUID /etc/auto.net", "failed": true, "rc": 257}

msg: Traceback (most recent call last):

  File "/root/.ansible/tmp/ansible-tmp-1461943252.91-163803999777791/setup", line 1668, in run_command

    cmd = subprocess.Popen(args, **kwargs)

  File "/usr/lib/python2.7/subprocess.py", line 679, in __init__

    errread, errwrite)

  File "/usr/lib/python2.7/subprocess.py", line 1153, in _execute_child

    self.pid = os.fork()

  File "/root/.ansible/tmp/ansible-tmp-1461943252.91-163803999777791/setup", line 1823, in _handle_timeout

    raise TimeoutError(error_message)

TimeoutError: Timer expired

When the fact gather task does work:

{

    "device": "/etc/auto.net",

    "fstype": "autofs",

    "mount": "/mnt/net",

    "options": "rw,relatime,fd=27,pgrp=27931,timeout=300,minproto=5,maxproto=5,indirect",

    "size_available": 0,

    "size_total": 0,

    "uuid": "NA"

}

Run manually

> lsblk -ln --output UUID /etc/auto.net

lsblk: /etc/auto.net: not a block device

Relevant bugs:

https://github.com/ansible/ansible/issues/11005

https://github.com/ansible/ansible/issues/6891

Has anyone come across this issue?  Alternatively, is there a way to disable certain facts from being gathered?

Thanks.