Latency-optimized geo-partitioning

Improve the latency of Multi-Active global applications using geo partitioning

For multi-active applications that need to be run in multiple regions, you can opt to partition the data per region and place the replicas in nearby regions. This ensures very low latency for both reads and writes for the local partitions, providing a seamless user experience for users close to their partitions.

Application instances are active in all regions, do consistent reads, and operate on a subset of data.

Overview

Setup

To set up a local universe, refer to Set up a local YugabyteDB universe.

Setup

To set up a cluster, refer to Set up a YugabyteDB Managed cluster.

Setup

To set up a universe, refer to Set up a YugabyteDB Anywhere universe.

Suppose you want to serve users both in the East and West regions of the US with reduced latency. Set up a cluster with a replication factor of 3 and leaders in 2 regions, us-west-1 and us-east-1, and place the replicas in nearby regions, us-west-2 and us-east-2. This provides low read and write latencies for the 2 partitions.

RF3 cluster spanning 2 regions

Partition your data

For the purpose of this example, create a basic table of users that you are going to partition by the geo field.

CREATE TABLE users (
    id INTEGER NOT NULL,
    geo VARCHAR,
) PARTITION BY LIST (geo);

Partition your data for east and west users. This ensures that the application in us-west will operate on west partition and the application in us-east will operate on the east partition.

The tablespace definitions are discussed in the next section. 
--  West partition table
CREATE TABLE us_west PARTITION OF users (
   id, geo, PRIMARY KEY (id HASH, geo)
) FOR VALUES IN ('west') TABLESPACE west;

--  East partition table
CREATE TABLE us_east PARTITION OF users (
   id, geo, PRIMARY KEY (id HASH, geo)
) FOR VALUES IN ('east') TABLESPACE east;

Partition your database

Replica placement

Configure your west partition leader preference to place the leader in us-west-1, one replica in us-west-2 (nearby region), and the other replica in us-east-2. Placing one replica of west data in the east has the advantage of enabling follower reads for applications in the east if needed.

--  tablespace for west data
CREATE TABLESPACE west WITH (
    replica_placement='{"num_replicas": 3,
    "placement_blocks":[
        {"cloud":"aws","region":"us-west-1","zone":"us-west-1a","min_num_replicas":1,"leader_preference":1},
        {"cloud":"aws","region":"us-west-2","zone":"us-west-2a","min_num_replicas":1,"leader_preference":2},
        {"cloud":"aws","region":"us-east-2","zone":"us-east-2b","min_num_replicas":1}
        ]}'
);

Place west replicas

Similarly, set up your east partitions in us-east-1, us-east-2, and us-west-2.

--  tablespace for east data
CREATE TABLESPACE east WITH (
    replica_placement='{"num_replicas": 3,
    "placement_blocks":[
        {"cloud":"aws","region":"us-east-1","zone":"us-east-1a","min_num_replicas":1,"leader_preference":1},
        {"cloud":"aws","region":"us-east-2","zone":"us-east-2a","min_num_replicas":1,"leader_preference":2},
        {"cloud":"aws","region":"us-west-2","zone":"us-west-2b","min_num_replicas":1}
        ]}'
);

Place east replicas

Low latency

Consider the West application. As you have placed the west partition's leader in us-west-1, it has a low read latency of 2 ms. As this partition has a replica in a nearby region, the write latency is also low (less than 10 ms).

West application

Similarly, the East application also has low read and write latencies.

East application

Failover

When any of the regions hosting one of the partition leaders fails, the partition follower in a nearby region would immediately be promoted to leader and the application can continue without any data loss.

Failover

This pattern helps applications running in different regions to have low read and write latency, as they are reading and writing data to nearby partitions. Note that the access latency increases a little to 10 ms and the write latency increases a lot more to 60 ms. This is because the writes have to be replicated to the follower in us-east-2, which is 60 ms away.

Learn more