What are your needs for persistent storage with Docker?

In a few days you will hear about a new version for our Docker Plug-in for InfoScale where we take advantage of new InfoScale 7.1 capabilities to provide quality of service and avoid the noisy neighbour problem. That means you will no longer have to worry about those uncontrolled applications that suddenly start affecting the performance of others. With 7.1 and the new integration, when Docker is creating a volume we can limit the maximum number of IOs per second it will be serving. It will be as easy as running:

docker volume create –d veritas –-name <volname> –o maxiops=10000

Now we would like to let you know other integrations we have in our pipeline and we would like to hear from you which ones are the most needed, so you can satisfy your needs first. Please click on the link to fill a small survey.

This is a description of the different things we are working on and we need your feedback:

I/O Acceleration

This feature will allow the Docker user to specify what volumes will need a higher I/O bandwidth and/or smaller latency so the data will be automatically cached on local SSDs in the hosts. When a container is reading data, it will be cached on local SSDs, so the next reads will be served from the SSDs. The only thing the Docker user will have to do is to specify that the volume needs to be associated with a cache by enabling iocache flag:

docker volume create –d veritas –-name <volname> –o iocache=on

Snapshots for local clones

Allow the Docker users to take snapshots of their persistent storage and make it available in any other hosts. This will allow the Docker user to work with different copies of the same data within the same cluster, bringing up containers that will be using that copy. We just need to specify the name of the new volume and what the name of the existing volume is:

docker volume create –d veritas –-name <volname> –o sourcevol=<volname>

Snapshots for backup/restore

Allow the Docker user to take snapshots that can be used for backup and restore purposes. To take a snapshot the user should only trigger this command:

docker volume snapshot –d veritas <volname>

To restore from one specify snapshot copy:

docker volume restore -d veritas <volname> <snap–number>

And to list the snapshots that are available:

docker volume snaplist –d veritas

Remote-clones

Take the persistent data copy of a container and make it available in any other cluster across any distance. This allows Docker users to work with different copies of the same data across different clusters.

On the local cluster we make the volume available:

docker volume export  -d veritas <volname>

On the remote cluster we just use that copy:

docker volume create –d veritas –-name <volname> –o remote_sourcevol=<volname>

Policy Management

Ability to create policies like gold, bronze, etc, so the Docker user can just point to them when creating containers:

docker volume create –d veritas –name <volname> –o policy=<class>

Integration with Kubernetes

Allow Kubernetes to run on top of an InfoScale Cluster

Integration with Docker SWARM

Allow Docker SWARM to run on top of an InfoScale Cluster

Graphical User Interface

Understand in a graphical view how the cluster is performing, how each container is using storage, where the storage is located and how IOs are balanced across the infrastructure

Encryption

Being able to encrypt a volume when created from the Docker CLI.

docker volume create –d veritas –-name <volname> –o encrypt=on

Please fill our survey and give us your feedback:

http://goo.gl/forms/Af8CMsvVlI2uKQb02

1 Comment

This is Cool! The rich features already available with Veritas CFS would be integrated with the Docker eco-system.