1

We have MySQL Multi-Master replication setup between two servers ( A & B ). It has been working for a few years now and life is good.

The management tool in our office connects directly to Server A to perform updates and selects. As you can imagine this is not the fastest, especially because our office internet connection is no good.

So I would like to replicate the whole database on our office server but do not want to do this...

A -> B -> C -> A

because C will be a bottleneck for replication between B & A

Is their a way to

A -> B -> A AND C -> A -> C

I only see examples for creating loops of servers in MySQL. We can not migrate to a different database at this stage, nor immediately abandon the existing replication between A & B.

2 Answers 2

2

As Sandor said, this cannot be done with standard MySQL replication. It can be done if you use a third-party / external replication mechanism.

A few products that I'm aware of:

Continuent Tungsten Replicator

SymmetricDS

Both are open source and freely available.

Another option would be to migrate to a MySQL cluster.

Some options:

MySQL Carrier Grade Edition

Percona XtraDB Cluster

If you decide to investigate the cluster route the terms multi-site replication and geographical replication will help you find what you're looking for.

0

No. That would require A to be slave to both B and C, which is not possible. A-›B->C->A not recommended either.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .