Fixed bug: If one ping fails, the subsequent analysis report that the ping is down.
- The reason was that in HostStatusAnalyzer, analyze_ping_failed_count() was
called from within increase_ping_failed_count(), thus only when the ping
failed the ExceededPingFailedLimit was updated, to false. There were no path
where ExceededPingFailedLimit was updated to true.
Now, the analyze_ping_failed_count() is called everytime we call
update_ping_statistics(), which updates ExceededPingFailedLimit everytime.