Bug 1170040

Description of problem:
For scaled application, if both head gear and scaled gears are in idle. the gear can be unidled whatever by accessing app's url or runing "oo-admin-ctl-gears unidlegear", but the time is too long ( about 4 mins), an error message  also be reported in platform.log

Version-Release number of selected component (if applicable):
OSE 2.2-2014-12-02.3/

How reproducible:
always

Steps to Reproduce:
1. create scaled applications .
2. idle all gears
3. access this application (curl http://appdns/) or run "oo-admin-ctl-gears unidlegear <app_uuid>",
4. check platform.log on node1(where head gear located.)

Actual results:
Step 4: it took about 4 minutes to unidle this application.

# time oo-admin-ctl-gears unidlegear domain-nonscalingapp-1
Unidling gear domain-nonscalingapp-1 ... [  OK ]
real    0m2.270s
user    0m0.716s
sys    0m0.150s

# time oo-admin-ctl-gears unidlegear domain-scalingapp-1
Unidling gear domain-scalingapp-1 ... [  OK ]
real    3m56.725s
user    0m0.822s
sys    0m0.163s

Step 5: there is error as below:
December 03 12:51:51 ERROR Failed to start application 547f674de5fed5b28b000038 during unidle: {"api_version":1.6,"data":null,"messages":[{"exit_code":-1,"field":null,"index":null,"severity":"debug","text":"Shell command '/sbin/runuser -s /bin/sh 547f674de5fed5b28b000038 -c \"exec /usr/bin/runcon 'unconfined_u:system_r:openshift_t:s0:c3,c479' /bin/sh -c \\\"set -e; /var/lib/openshift/547f674de5fed5b28b000038/haproxy/bin/control start \\\"\"' exceeded timeout of 234"},{"exit_code":-1,"field":null,"index":null,"severity":"error","text":"A gear start did not complete on 1 gear. Please try again and contact support if the issue persists."}],"status":"internal_server_error","supported_api_versions":[1.0,1.1,1.2,1.3,1.4,1.5,1.6,1.7],"type":null,"version":"1.6"}

Expected results:
1) it should takes little time to unidle gears
2) no error was show during unidle gear