Distributed Command Bus Rebalancing on Kubernetes after Redeployment Issue

Imported from Google Groups
1

Hello,

We are having a problem with the Distributed Command Bus. We are using Axon Framework + Springboot running on Kubernetes. We have configure the Spring Cloud Kubernetes integration for service discovery.

It is working well, expect when we rollout a new version of our application to Kubernetes. In this case, the topology of the members is changed and we noticed it took about 10 minutes to rebalance. In this time, we were not able to process any commands.

Is this expected behaviour? Can we do something to resolve/avoid this issue?

Kind Regards,
Georgios Chinis & Yuriy Bravenko

2

Hi Georgios.

the discovery mechanism uses heartbeats to periodically sync data between available nodes. Depending on the duration of these heartbeats, it may take a while for nodes to be completely in sync. 10 minutes does sound like a long time, though. Generally, it’s in the order of magnitude of 30 seconds to 1 minute.

Note that AxonServer is significantly simpler and faster in these situations. It will seamlessly distribute your commands, queries, and events across deployed instances. Have you considered using it?

Cheers,