Simple and powerful
high performance state replication without
the complexity
Replicated Hash Table (RHT) is an in-memory
hash table data structure replicated on
multiple participating computer systems.
Consistency between each table replica
is maintained at all times. RHT automatically
handles network partitions, server crashes,
and network re-merges between the various
replica.
To the programmer, RHT exposes a simple
and intuitive common interface to the
standard key-value hash table data structure.
|
|
- Simple to use and
deploy.
- Takes the complexity
out of multiple machine
state replication.
- Increase query
throughput linearly
with the number of
replicas.
- Reads are handled
locally.
- Writes/Updates
are replicated with
guaranteed consistency.
- Supports thousands
of updates per second
in typical setups.
- Synchronized state
is maintained in the
event of server failures/recoveries
and network partitions/remerges.
- Snapshots of the
hash table can taken
at any time.
- Supports iterators
as well as Get/Put
calls.
- Scales to dozens
of replicas
- C/C++ interface
- Thread safe
- Cross platform (Linux,
Windows, Mac, and
others)
|
|
|
|
|
|
