High Availability - Network File System (NFS) between EC2

I have the following scenario:

• 11 EC2 servers with Docker and about 35 containers (Web application) on each host;
• An NFS server that shares files between hosts. Containers needs to access these files.
• Rancher Master Server to orchestration these containers.

Well, the question is: We have a single point of failure that is NFS Server. I need suggestion to avoid this disaster and so...:

1. HA NFS with DRBD and HeartBeat;
2. EFS is unavailable in my region;
3. "GlusterFS was not a good solution in the case that the web servers were writing small files (meaning small number of kilobytes) often that change a lot e.g. session.xml being read, updated re-saved, read, updated, re-saved etc."
4. S3 with S3FS-FUSE...but Amazon does not recommend S3 for a File System and don't have a native tool to mount .

I read about Aws Storage Gateway but this only makes sense in Hybrid Scenarios (Cloud + On-Premise)...in my case, just interest in Cloud.

I see no other solution than #1. Does anyone have another point of view ? Any Suggestion?

Thanks.

Edited by: markketing on Apr 10, 2019 11:31 AM