As container adoption ramps up, some of the typical operational issues with stateful workloads continue to exist. For example, containerized MySql instances might have to resize the storage (volume), or need to spin up additional MySql instances based on demand. DevOps teams are keen to realize some of the operational advantages with containers and Kubernetes for stateful workloads as well.
Datera solves these key operational challenges with flexibility and programmability of the data managed.
Datera Elastic DataFabric (EDF) enables developers and DevOps with application-driven flexibility in container environments.
The key benefits Datera provides for dynamic stateful workloads such as MySql, Redis, Cassandra, and so on are:
- 1. Online expansion of volume size – Without any disruption to applications, volume size can be expanded
- 2. Progressive configuration changes on volume replica count – Based on the application availability and IO durability requirements, the volume replica count can be reconfigured
- 3. Programmatic storage provisioning at container speeds and scale – Ability to provision volumes are scaled through the volume plugin, or through the API
- 4. Workload controls at container granularity – With per-volume QoS controls for IOPS and bandwidth, Datera provides per-container controls
- 5. Automation of storage scaling through pre-defined templates – AppTemplates to automate scaling of storage tracking container auto-scaling.
Persistent Storage Provisioning in Kubernetes
PersistentVolume (PV) can be statically or dynamically provisioned by the administrator. This is usually a cluster-wide resource. Datera implements PVs through flex volume-based volume plugins to present the Datera Volumes to the cluster. PVs manage the lifecycle of the volume independent of any individual pod that uses the PV.
PersistentVolumeClaim (PVC) is the desired storage resource request by a consumer. Pods consume node resources, and PVCs consume PV resources.
For the Datera SDK and Volume Plugin, please go to https://github.com/Datera