Baud is a multi-model storage system with reliability, scalability, and flexibility.
Field, Object, Class, Edge, Space
each object has a field as the 'partition key' and an internal 'Object ID' (OID) generated by the system.
each field can be indexed.
runs on JDOS/Kubernetes across servers, racks, cells and regions
space --> partition
partition ID range
raft
async
filtered
there are several kinds of nodes in a Baud cluster: router, toposerver, objectserver, blobserver, which all speak with HTTP+JSON
strong indexing and search capability
src --edge--> dst