-
Notifications
You must be signed in to change notification settings - Fork 8
Home
Welcome to the IginX wiki!
IginX is a database clustering system for multi-dimensional scaling of standalone time series databases through generalized sharding.
By encapsulating shard orchestration logic, IginX allows application code and time series data queries to remain agnostic to the distribution of data onto multiple shards. With IginX you can even split and merge shards in multiple dimensions as your needs grow, with an atomic cutover step that takes only a few seconds.
IginX is still at a premature phase, but you are highly encourage to try it and share us with your experience. Currently, IginX works best with IoTDB.
The following picture overview the deployment architecture of IginX. The deployment of an IginX-based distributed time series database typically consists of two layers plus a few metadata servers. The IginX layer manages and orchestrates the underlying TSDB layer, with metadata kept in the metadata servers. The coupling between layers and with each layer is loose, making the system highly flexible and easily scalable.
The following image depicts the major component framework of IginX. The design of IginX features a flexible data partitioning scheme, pluggable data distribution strategies and pluggable query processing policies.