Question
Has anybody deployed their own database cluster across data centers in a region?
- Posted on April 4, 2025
- High AvailabilityMariaDBNetworkingVirtual Private Network
Asked by David Nusbaum
I have been considered a MariaDB Cluster (MaxScale or Galera) with a node in NYC1, NYC2 and NYC3. With vpc peering this seems like a reasonable approach for a HA solution. But…
- With MaxScale write connection are going to frequently cross between data centers to hit the primary and read could cross between data centers as well.
- With Galera only the synchronous replication will cross data centers, but writes will block until the data is certified on all three servers.
The latency looks pretty good between datacenter, but I cannot replicate the full load or our solution in order to get a clear understanding of the performance implact. I have tested to make sure it works and performance for test transactions seems pretty good.
I’m just reaching out to see if anybody else has already attempted this and might have some knowledge to share.
This textbox defaults to using Markdown to format your answer.
You can type !ref in this text area to quickly search our full set of tutorials, documentation & marketplace offerings and insert the link!
These answers are provided by our Community. If you find them useful, show some love by clicking the heart. If you run into issues leave a comment, or add your own answer to help others.
Become a contributor for community
Get paid to write technical tutorials and select a tech-focused charity to receive a matching donation.
DigitalOcean Documentation
Full documentation for every DigitalOcean product.
Resources for startups and SMBs
The Wave has everything you need to know about building a business, from raising funding to marketing your product.
The developer cloud
Scale up as you grow — whether you're running one virtual machine or ten thousand.
Get started for free
Sign up and get $200 in credit for your first 60 days with DigitalOcean.*
*This promotional offer applies to new accounts only.
Hey David 👋
Really interesting setup, running a MariaDB cluster across NYC1, NYC2, and NYC3 with VPC peering definitely seems possible, but like you said, the big unknown is how it’ll behave under real-world load.
With MaxScale, there’s a good chance that write traffic will frequently cross DC boundaries, depending on how your routing and failover is handled. That could introduce some inconsistent latencies, especially during bursts or failover scenarios.
I’ve used Galera in the past, but only in the same region. It is a bit more predictable since it uses synchronous replication, but that also means every write has to be confirmed across all nodes. Even if latency looks fine in your tests, it’s hard to know for sure until you simulate real production-level traffic. And Galera can be sensitive to even small increases in round-trip time when the write volume goes up.
Honestly, I’ve seen setups like this work in some cases, particularly for read-heavy or moderate workloads, but they can become tricky to maintain as things scale or possible network hiccups.
If this is mission-critical and HA is the priority, it might be worth considering whether a managed database service could take that operational complexity off you. Like DigitalOcean’s managed databases with high availability:
Would love to hear how it goes if you decide to move forward, this is the kind of experiment others would definitely benefit from learning about.
- Bobby