Stolon vs CrunchyData for HA Postgresql

Question

I'm using helm charts with kubernetes and I'm trying to deploy High Availability Postgresql database and found this article mentioning three project that provide HA Postgresql:

1- Stolon

2- CrunchyData

3- Patroni

I didn't find helm chart example for Patroni, but I installed both helm charts for Stolon and Crunchy. However, I wasn't able to find the differences between them except in the structure, both use statefulset from kubernetes in order to deploy master service and a number of replicas(ReadOnly slaves),stolon has this proxy service that points to the master service which is like an extra layer, makes the stolon look a little more complex than crunchy.

But what I really need to know is:

1. How the replicas will synchronize with the master? synchronous or
   asynchronous.
2. Whats the failover mechanism provided if any?
3. persistence volumes why is there a volume per pod? How is this related to the replication?

Answer 1

I ended up using stolon for its automatic leader election and fail over handling. Crunchy Data offers small building blocks so there is a lot of work to do to put them all together while stolon has a ready helm chart. But I didn't try patroni since stolon worked just fine.