mcs-savebrm.py overwrites Extent Map files multiple times with shared(non-S3) storage setup

When MCS cluster is shutdown mcs-savebrm.py is called to save extent map. The original intention was to save it only on primary node but mcs-savebrm.py doesn't detect a primary in a shared(non-S3) storage setup. This effectively allows cluster nodes to overwrite extent map files multiple times during shutdown and in some cases causes save_brm binary to stuck.
The suggested solution is to start detecting primary in mcs-savebrm.py for any type of the cluster or storage. There are two detection mechanisms implemented in mcs-savebrm.py:

• using CMAPI
• search for DBRMController.IPAddr in the list of local IPv4 addresses (if CMAPI is unavailable)