Add option to disable overriding the status basing on the largest process memory in pmp-check-unix-memory

[Roman Vynar]

By default, pmp-check-unix-memory not just checks the free memory with the warning and critical thresholds but also compares them with the largest process memory usage.

So in case I have a large process consuming most of the memory I see something like this:

[root@centos6 ~]# /opt/percona-nagios-plugins/pmp-check-unix-memory -w 90 -c 95
WARN OK Memory 85% used. Largest process: mysqld (24579) = 15432.75MB (90%)
[root@centos6 ~]# /opt/percona-nagios-plugins/pmp-check-unix-memory -w 90 -c 95
CRIT WARN Memory 93% used. Largest process: mysqld (7008) = 16846.75MB (99%)

As you can see the specified thresholds trigger one status first and then the another final status appended according to the comparison with the largest process memory.
I think I should be able to disable this behaviour because on the dedicated boxes I could have MySQL tuned with innodb buffer pool to 80% of total memory and it will end up with status problems on Nagios even there are no ones actually.

I should be able to perform:
[root@centos6 ~]# /opt/percona-nagios-plugins/pmp-check-unix-memory -w 90 -c 95 -<DISABLE LARGE PROCESS CHECK>
OK Memory 85% used. Largest process: mysqld (24579) = 15432.75MB (90%)

We should be consistent with the specified/default thresholds and do not depend from the largest process memory consumption optionally.