November 2009: Offering media services in over 30 languages, the internationally acclaimed Deutsche Welle was seeking to monitor their equally diverse IT. Large and distributed across two offices in Bonn and Berlin, the broadcaster needed to consolidate their disparate monitoring activities into one enterprise system. A central overview was essential, as was exceptionally high availability. Deutsche Welle offered NETWAYS the opportunity to suggest a solution- and an innovative one was proudly delivered.
Since monitoring needed to be self sufficient at each location while being centrally accessible, the consulting team designed a multiple redundant system. Two monitoring clusters were implemented in each city and integrated by a single MySQL cluster. In a master/master failover, each database was also physically located in their respective city. This amounted to two physical Nagios servers and one database at each office, which could operate independently if ever a break in connection were to occur. Furthermore, two physical Nagios servers at each location gave additional guarantee of constant availability. By using the MySQL cluster as a kind of interchange hub, monitoring was centralised without compromising operational independence.
DeutscheWelle_distributed_monitoring
Upon implementation, an array of Nagios addons and plugins were also employed. From NSClient++, SNMPTT and NagVis to EventDB, NagiosGrapher and Business Process Addon, Deutsche Welle was fully equipped by the end of the project. All that is left is a few days training to equip the staff with the skills for a similar operational independence.