Introduction and overview of main principles of TimeBase replication.
Tags: replication

Replication allows duplicating data from one stream (source) to another stream (replica) and helps to tackle important issues such as read/write performance, scalability, data backup/restore, fail-over protection, data distribution, and access control to name a few. It can be configured to copy all the existing streams or selected stream(s). The minimum replication unit is symbol. You can replicate data between two ore more TimeBase instances or copy data to external data storage and processing services such as S3, ClickHouse, Timescale, Kafka.

Synchronous Replication

TimeBase can be used as a persistent message broker which allows for Live replication (may be as well called a streaming or synchronous replication). Live replication maintains an open connection between the source and the replica(s) streams, where all changes are applied to replica(s) right away (with ms latencies).

Asynchronous Replication

The source and the replica do not have to be connected all the time. You can configure a scheduled or so-called batch replication to copy data on a specific time or, for example, during off hours. When replication is run in the batch mode, it downloads all the available changes from the source stream, closes connection, and stops.