Vivek
unread,Mar 23, 2018, 3:31:19 PM3/23/18Sign in to reply to author
Sign in to forward
You do not have permission to delete messages in this group
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to Kubernetes user discussion and Q&A
Hey team,
I have an application container that I wish to run on Kubernetes in an AWS environment, backed by EBS volumes.
I’m currently thinking that the application will be created by a StatefulSet with a single PersistentVolume attached (EBS underlying).
I had originally thought that if the AWS availability zone went down, Kubernetes would automatically spin up the node again in a different AZ, attaching to the same volume. That said, I then realized that EBS volumes only exist within the same AZ as the instance to which they are attached.
From the AWS documentation, It seems the proper way to duplicate an EBS across AZs is to create a volume snapshot and then spin up another volume and attach that new volume to a new instance in another AZ.
That said, I’m not sure if StatefulSets in Kubernetes have the necessary logic to accomplish this. Questions:
- Is there a way that I can specify for StatefulSets to create a persistent volume from a snapshot in the event that the original volume does not exist for a pod?
- Alternatively: does anyone see an easier architecture that I might opt for instead, given these application constraints?
Disclaimer: I’m fairly new to Kubernetes, so I’m sure that I have made some obvious logic errors!
Regards,
Vivek