(WIP) Use dynamic IPv4 network pools for deploy clusters (#18)

*     Use dynamic IPv4 network pools for deploy clusters

    - Use 'config' object as an additional jinja option to load templates
    - Use 'address_pools' from 'config' object in
      virtual-mcp11-ovs and virtual-mcp11-dvr
    - Fill 'address_pools' in 'config' object from fuel-devops manager

*     Use dynamic IPv4 network pools for deploy clusters

    - Use 'config' object as an additional jinja option to load templates
    - Use 'address_pools' from 'config' object in
      virtual-mcp11-ovs and virtual-mcp11-dvr
    - Fill 'address_pools' in 'config' object from fuel-devops manager

* return removed 'test_install_k8s'

* Use 'sed' to replace IP addresses in the models to pre-generated

also, two new environment variables avaliable for mcp11-ovs/dvr:
- SALT_MODELS_REPOSITORY
- SALT_MODELS_COMMIT

* remove temporary env_config.py.bak
diff --git a/tcp_tests/settings_oslo.py b/tcp_tests/settings_oslo.py
index 8190720..5bdf6b0 100644
--- a/tcp_tests/settings_oslo.py
+++ b/tcp_tests/settings_oslo.py
@@ -77,6 +77,12 @@
            default=[]),
     ct.Cfg('lvm', ct.JSONDict(),
            help="LVM settings for Underlay", default={}),
+    ct.Cfg('address_pools', ct.JSONDict(),
+           help="""Address pools (dynamically) allocated for the environment.
+                   May be used to determine CIDR for a specific network from
+                   tests or during the deployment process.
+                   {'pool_name1': '<cidr>', 'pool_name2': '<cidr>', ...}""",
+           default={}),
 ]