Of the three primary Linux vendors (Canonical, Novell, and Red Hat), Canonical and Red Hat have made the biggest splashes in the cloud computing market. Canonical’s focus appears to be simple partnerships and bundling software, rather than the comprehensive enterprise products offered by Red Hat. At its 2010 Summit, Red Hat provided a complete and separate track of cloud sessions that introduced its family of cloud products and services, along with its cloud strategy. While Red Hat provides an abundance of information about its cloud offerings, it’s not always clear how they fit together.
The overarching strategy behind Red Hat’s cloud offerings is to provide a consistent environment that allows you to run your workloads in your enterprise data center (fully or partially virtualized, with or without a private cloud) or in a public cloud. This consistency extends all the way through licensing.
For example, if you exhaust capacity in your data center, Red Hat software, specifically MRG Grid, can automatically schedule workloads on virtual machines in the Amazon public cloud. Of course, you get to specify which workloads that you are willing to allow to be run outside your data center. MRG Grid is designed to schedule various types of computing resources, including virtual machines across private and public clouds.
Building a cloud stack with Red Hat
Red Hat’s cloud offerings do not lock you into its cloud stack. Instead, they allow you to build a cloud stack that can be used alone or with components from third-party vendors. This approach differs significantly from VMware’s approach. When VMware publishes application programming interfaces (APIs), it often keeps the best features for itself. An entire VMware stack offers quality performance levels, but when you introduce third-party components into the VMware stack, performance often suffers.
Red Hat allows you to use as many tools as you want. You can plug in your own homegrown tools or third-party management tools with no performance or functionality penalty; the open source model offers third parties access to complete sets of APIs.
Most Fortune 1000 enterprises are Red Hat customers, and Red Hat is looking to expand its cloud technologies within these enterprises. If an enterprise has VMware, then instead of urging the enterprise to tear out VMware and replace it with Red Hat cloud software, the enterprise can keep VMware and expand with Red Hat. If you happen to be an enterprise using VMware, you can use Red Hat’s migration tools to move workloads from VMware ESX to Red Hat KVM and back again. Red Hat’s virt v2v tool is specifically designed to automate the conversion of VMware- or Xen-based virtual servers to KVM-based virtual servers and vice versa.
If you have a Red Hat private cloud and you decide that you want to create a virtual machine (VM) on another cloud, such as a Hyper-V, and deploy a workload there, the Red Hat Deltacloud tool provides you with the appropriate management interface.
Deltacloud permits you to enable and manage a heterogeneous cloud virtualization infrastructure, including Amazon EC2, GoGrid, OpenNebula and Rackspace. Instead of having a management console for VMs based on ESX and a management console for VMs based on Hyper-V, your VMs can now be managed from a single console. Deltacloud has recently been transitioned to the Apache Foundation as an incubator project to ensure that it is viewed as a full-blown open source community project.
Managing cloud workloads with Red Hat
While Red Hat’s MRG Grid can schedule workloads across clouds, Red Hat Enterprise Virtualization Manager (RHEV-M) manages and schedules workloads within a single cloud. The open source KVM hypervisor, rather than Xen, is the basis for Red Hat’s RHEV offering. KVM is an upstream hypervisor, which means that it ships with the Linux kernel and takes advantage of Linux kernel updates automatically.
Latest posts by cloudtweaks (see all)
- Companies Turning To The Cloud To Meet Increased Customer Expectations - March 10, 2014
- Cloud Infographic: Disaster Recovery - March 9, 2014
- Facebook Finding Friends In The Node Pole - March 7, 2014