Container storage: 5 key issues it’s worthwhile to know
Most enterprises now run functions in containers, and they also should take note of how they retailer and handle knowledge for containerised functions.
The Nutanix enterprise cloud index, compiled for the cloud software program provider by VansonBourne and revealed earlier this 12 months, discovered 54% of companies have containerised all their functions and as many as 98% run at the very least one occasion of Kubernetes.
This, nevertheless, poses plenty of challenges for IT architects with regards to knowledge storage. Containers have been designed to be ephemeral, or short-term, in nature. This works nicely sufficient for microservices. However mainstream enterprise functions must course of and retain knowledge. This has required builders to adapt container know-how to assist persistent storage.
Containers convey a spread of benefits to enterprises. Containerised functions run independently of the host working system, making them extremely transportable. This helps companies that run functions in hybrid or multicloud environments.
Containers are additionally “mild”, demanding fewer sources, particularly storage, than typical digital machines. Containers are extra environment friendly, and spin up in seconds slightly than minutes.
And, whereas not all containerised functions are microservices, the effectivity of containers lends itself to working them and permits the development of advanced functions out of small, reusable and environment friendly components.
How do containers and storage join?
The primary era of containers have been designed to be stateless. This had benefits in pace of deployment and effectivity. However stateless, or impermanent, functions can not retailer knowledge past the lifetime of the container.
Stateless functions work in some conditions, reminiscent of net companies or microservices that don’t must retailer and entry knowledge on an ongoing foundation. However that capacity to deal with knowledge is central to many, if not most, enterprise functions.
Consequently, container know-how has tailored by including persistent storage. Persistent storage sits outdoors the container, and might be on on-premise or cloud {hardware}, as file, block or object storage.
The container orchestration layer manages persistent storage. Within the case of Kubernetes – the most typical container orchestration system – knowledge is saved in persistent volumes (PVs) and provisioned through persistent quantity claims (PVCs) which can be transportable and might transfer with the container.
PVs are impartial of any pod and should not transportable throughout Kubernetes clusters. Each, nevertheless, serve to decouple the container and the storage, in order that “typical” storage works with containerised functions.
The problem for IT groups, nevertheless, is that that is removed from plug-and-play. The containerised functions, the orchestration layer and the storage all must work collectively seamlessly to permit an enterprise utility to work.
How does CSI assist with knowledge storage for containers?
To simplify and standardise how containers hook up with storage, the business has developed the container storage interface (CSI) and container-native storage.
CSI works with cloud, on-premise and hybrid storage, and throughout file, block and object storage. This permits builders to tailor their storage to their workloads.
CSI is a set of requirements that permit storage suppliers to attach their know-how to Kubernetes. At present, there are greater than 100 completely different CSI drivers accessible, for normal and software-defined storage.
CSI continues to evolve, including assist for extra storage codecs and extra suppliers. One additional benefit of CSI is that it helps IT groups constantly handle storage, even throughout a number of suppliers’ infrastructure.
What’s container-native storage?
Container-native storage, for its half, is software-defined storage working contained in the container, on Kubernetes. Container-native storage gives the prospect of solely allocating storage to the container when the container wants it, making it extra versatile than different types of storage.
Suppliers providing container-native storage embrace Crimson Hat’s OpenShift Information Basis (ODF), Pure’s Portworx and Nutanix’s Unified Storage.
In line with business analyst Gartner, 95% of world organisations could have containerised functions in manufacturing by 2029.
The cloud-native storage market, nevertheless, is much less mature. Trade estimates put take-up for Portworx and Crimson Hat ODF, mixed, at below 30% of the market, though analysts count on the market to greater than double by the top of this decade. This implies that, for now, enterprises are sticking with CSI.
How do storage suppliers assist container storage and backup?
Suppliers are working to make container storage simpler to handle and higher capable of work throughout a spread of storage applied sciences. That is all of the extra essential for enterprises that function hybrid clouds. Some companies need to preserve storage in-house or in non-public clouds, however nonetheless need to make the most of cloud-native and containerised functions.
Consequently, suppliers together with Dell EMC, HPE, Hitachi Vantara, IBM, NetApp and Pure have every labored to enhance assist for containers. The hyperscale cloud suppliers have additionally continued so as to add to their container assist.
These applied sciences at the moment are pretty mature and, as such, ought to permit enterprises to run containerised functions with persistent storage with out altering {hardware} or cloud storage provision.
On-premise and cloud suppliers have additionally added higher assist for backup and restoration of containerised environments. Strong backup is a vital characteristic for enterprise manufacturing deployments.
What container administration merchandise do suppliers provide?
The problem stays, nevertheless, to strip out extra of the complexity round persistent storage and containers.
Instruments reminiscent of Pure’s Portworx and NetApp’s Trident goal to simplify provisioning of Kubernetes functions, in addition to enhance portability and safety.
NetApp’s Trident is open supply, freed from cost and makes use of CSI. It helps automated provisioning for NetApp OnTap storage as PVs for Kubernetes. Trident additionally gives knowledge administration, knowledge safety, and catastrophe restoration and enterprise continuity for container environments.
Portworx additionally gives automated knowledge companies and policy-driven administration. Portworx additionally makes use of CSI, and swimming pools underlying storage right into a single knowledge material. That is then shared throughout clusters.
Pure factors out that Portworx gives a constant mannequin for storage throughout cloud, hybrid and on-premise storage, with “cloud-like agility and responsiveness” for on-premise environments.
Pure lately built-in its Fusion clever management aircraft into Portworx, and added a synthetic intelligence (AI) co-pilot, which it says can monitor Kubernetes clusters at scale. Portworx additionally integrates backup and restoration and automatic capability administration into its platform.
These developments ought to all make it simpler for builders to create containerised functions that want persistent storage, but additionally take away a few of the overheads of storage and knowledge administration, in addition to catastrophe restoration, as soon as containerised software program is working in manufacturing.
