Hello Team,
This is just a thought to preserve the data files generated by AxonIQ-SE, where in replication is not supported. So was thinking if the underlying file generated by axon server is replicated by underlying file system then eventhough the axon sever is crashed or the AWS AZ is failed on which the node is running, the data would be intact as its replicated across AZ’s.
Does these files generated out of axon server, be able to be stored onto AWS EFS for them to be replicated across AZ ?
Hello @mnkartik,
The short answer to this is “yes,” but there are definitely some gotcha’s to take into account.
Note that Axon Server (both SE and EE) is, in most cases, high on I/O.
Sure, it depends on what your connected applications are doing, but likely you’re going to dispatch a lot of commands, event publication and reading, and query sending.
With that in mind, you cannot come by with the default EFS settings, as it limits your IOPS. Especially the Event Store I/O will suffer from this.
You can mitigate this by paying a premium, of course, but whether that’s beneficial compared to simply going for Axon Server EE is a different topic.
So in short, I believe you can manage with AWS EFS, given that you very diligently monitor the space you’re using and the I/O. Whether that’s worth the effort, is up to you and your team.