Too much DNS load on pod readiness with kubernetes

[jlamaille]

🐛 Bug Report

We use zalenium with kubernetes, the readiness status causes a heavy load on our DNS.
Indeed the use of the method "getent" causes calls to the DNS which tries to solve urls of the type:
disulfur-40000-c757p.google.internal
disulfur-40000-c757p.x.c.f ..... internal
etc.
A call made by the readiness status generates at least 4 calls on the DNS times the number pods with a short frequency, the load is very important.

To Reproduce

Deploy a Zalenium infrastructure with a service account in a kubernetes environment and a corporate dns.
There are load peaks on the DNS due to the readiness of Selenium nodes.

Expected behavior

No load peaks on the DNS.

Environment

OS: Linux
Zalenium Image Version(s): last version
Version Docker : 18.09
Version Kubernetes : 1.13.7-gke.8

[jlamaille]

I made a PR to fix that issue : #1048

[diemol]

Closed via #1048