YugabyteDB can continuously serve requests in the event of planned or unplanned outages, such as system upgrades and outages related to a node, availability zone, or region.
YugabyteDB provides high availability (HA) by replicating data across fault domains. If a fault domain experiences a failure, an active replica is ready to take over as a new leader in a matter of seconds after the failure of the current leader and serve requests.
This is reflected in both the recovery point objective (RPO) and recovery time objective (RTO) for YugabyteDB universes:
- The RPO for the tablets in a YugabyteDB universe is 0, meaning no data is lost in the failover to another fault domain.
- The RTO for a zone outage is approximately 3 seconds, which is the time window for completing the failover and becoming operational out of the remaining fault domains.
YugabyteDB also provides HA of transactions by replicating the uncommitted values, also known as provisional records, across the fault domains.
The benefits of continuous availability extend to performing maintenance and database upgrades. You can maintain and upgrade your universe to a newer version of YugabyteDB by performing a rolling upgrade; that is, stopping each node, upgrading the software, and restarting the node, with zero downtime for the universe as a whole.
For more information, see the following: