Deploying on a High Availability Architecture

0
Hello,   We would like to have our Mendix application to be deployed on a high availability architecture, and based on our research we found that there are some limitations with Mendix few years ago. We are thinking if these limitations are overcome in the newer versions, so we have the following inquiries: Is it possible to have 2 App Servers (having the same app version in both) connected to 1 DB Server at the same time? where if one App server is down the second App server will be up and running. Is it possible to have 2 App Servers (having the same app version in both) connected to 2 DB Servers (MSSQL DB to be synced) at the same time? where if one DB is down the second DB will be running and app servers should automatically reconnect to the second DB. What is the best way to achieve a high availability architecture kind of servers?   Thank you! Haitham
asked
1 answers
0

Hi Haitham,

 

The App servers, when nodes on a Kubernetes cluster, can connect to the same DB. See here for details: https://github.com/aws-ia/terraform-aws-mendix-private-cloud/tree/main. It's written for AWS, but you will get the point for Azure and on-prem, too. Another good source is here: https://docs.mendix.com/refguide/clustered-mendix-runtime/ including the limitations https://docs.mendix.com/refguide/clustered-mendix-runtime/

 

 For the database, I'm looking at https://www.yugabyte.com/. We are planning to test Yugabyte in the next weeks. From an architecture point of view, it should work as it is 100% PostgreSQL. I haven't experience with an active-active MSSQL DB. For Mendix, it is important to see the database on one IP address regardless of what node is servicing.

 

Go Make It

 

 

 

 

 

 

 

 

 

answered