Understanding Topology
Provides an overview of how to define cluster topology.
The MapR software uses node topology to determine the location of replicated copies of data. Node topology describes the locations of nodes in a cluster. You can define the cluster topology by specifying a topology for each node in the cluster. Use topology to group nodes by rack or switch, to provide a hint as to how data should be replicated to protect against data loss or unavailability because of a switch or rack failure.
In a topology, MapR
distributes container copies optimally among leaf nodes. For example, in a topology such as
europe/uk/london/DC2/room4/row22
, where row22
contains
multiple racks such as row22/rack1
, row22/rack2
,
row22/rack3
, and so on, MapR tries to ensure that all copies of the container do not end up on the same
rack (for example, rack1
). By setting each leaf value to correspond to a
physical rack, you can ensure that replicated data is distributed across racks to improve
fault tolerance.