Jelastic Hybrid Cloud with Multiple Regions

Note: This document is based on Jelastic version 3.3.

The basic idea of the Multi-region feature is fairly simple - to make the cloud hosting truly universal through giving an ability to aggregate various types of hardware, IaaS and 3rd-party cloud tools, within a single Jelastic installation. Herewith, all of the included services are orchestrated within a single infrastructure.

Despite the simple concept, such an approach brings the doubled efficiency through ensuring extra distribution possibilities for both hosting service providers/ISV companies and their customers: for the first group, it allows them to grow locally and/or conquer the remote market, while for the second one, it gives the impressive flexibility in application lifecycle management and smart organization of the dissemination policy.

General Overview

The necessity to use different hardware can be driven by different purposes, e.g. to vary the provided equipment based on the specific parameters, distant geographical location, access policies and so on. Thus, within the Multiple Regions feature implementation at the Jelastic Platform, we discern the following types of hardware formations:

  • Region - an individual hardware identity with its own subnet, domain and Shared Resolver(s). It can be located at the remote datacenter or represent the external 3rd-party service and still being managed by means of the common for all regions cluster orchestrator. Herewith, each region can contain multiple hardware node groups.

  • Hardware node (Hardnode) Group - a separate hardware set within the confines of a particular region with its own options, efficiency and rules for resources charging. The availability of each such set for users can be configured severally.

Generally, regions are used to unite different hardware node groups, that belong to the same datacenter, but vary in parameters. Herewith, at the end-users’ side this partition is shown simplistically - they operate directly with the set of hardnode groups, that are available for them (i.e. from all regions), while the regions’ partition itself remains implicit.

Such a division forms a clear, highly flexible and well-organized structure, which wipes out the borders between Public and Private Clouds for building a truly Hybrid Cloud. Here is a short list of benefits and opportunities you are able to receive with it:

  • expand the zone of influence through providing the corresponding services all over the globe with multiple hardware sets acquired
  • improve the service quality through broad geo distribution (as clients, that are physically located closer to the data center, get a significantly better response time)
  • ensure a real versatility for users through proposing to choose between the higher quality from one hand and some more affordable variants from the other, by means of a region’s hardware splitting into hardnode groups, with different capacity
  • special disaster recovery and safety ensurance through keeping data across multiple datacenters and services

Possible Scenarios

Beside the evident approach of distributing regions across data centers in different locations, such a model brings an ability to implement a great variety of different scenarios for your cloud - just to give you an idea of that, let’s consider several admissible solutions:

  • allocate a dedicated region or hardware node group for a particular type of user (e.g. for small/trial customers and/or enterprise clients separately) through the elementary hardware access permissions’ managing

  • ability to add cloud resources from external public clouds without superfluous efforts, due to a fairly plain integration

  • provide truly flexible hosting with the user-called environment migration enabled, which allows moving of projects among different regions/hardware node groups, depending on the current development stage

And now, let’s reveal in more detail how your desired Multi-Region structure can be implemented - the following guides will help you with that: