Cloud Bursting: Is It Ready For Enterprises?

Cloud bursting means workload moves from one cloud to another on the fly. With differentiation of public cloud and private cloud, you can have 4 different permutations: private to public, private to private, public to public, and public to private.

What people talk about the most is the case of private to public cloud. Think about the case in which an enterprise suddenly has more workloads than its infrastructure can handle. It’s very natural to move some of the workloads to one or more public clouds.

To achieve a successful cloud bursting, you will first need to move the workload. If it’s a virtual machine, just vMotion or simply copy the virtual machine files. After the workload up running in the new cloud, you need to route the traffic to it, for example using a load balancer. From users’ perspectives, they don’t feel they are using other clouds.

Overall I think cloud bursting is great in theory, but based on a hypothetical assumption about the enterprise workloads.

First, most enterprise infrastructure is under-utilized. As Carter Shanklin pointed out in his tweet, “What’s all this talk of cloudbursting? Aren’t datacenter utilization rates still generally below 10%? Figure out how to use the 90.” Well said! When your datacenter is not fully utilized, there is no point to burst outside.

Secondly, unlike Web 2.0 companies, enterprises don’t normally see erratic workload patterns. I’ve heard a Web site needs 1,000 servers over a weekend but just several during weekdays. In that scenario, you do get a lot of saving by bursting into a public cloud. For a typical enterprise, the types of workloads are mixed and they can offset each other from time to time. You can also schedule these workloads some times. Even occasionally you cannot, the increase will never be as dramatic as several times or hundreds of times as Internet companies. With this in mind, I do think Internet companies need cloud bursting. Or, why bother with bursting? Just live wholly in a public cloud!

In conclusion, I don’t think cloud bursting has arrived its prime time, maybe never. Instead of pursuing these cool projects, an enterprise should really focus on its private cloud. Also every business is different, so don’t be distracted by Web 2.0 companies.

Thoughts?

This entry was posted in Cloud Computing and tagged , . Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.

4 Comments

  1. Sebastian
    Posted August 16, 2011 at 8:38 am | Permalink

    I see it twofolds: cloudbursting and hybrid clouds. Agree with you that business case for cloudbursting if weak – for exactly the same reasons as you listed. However hybrid clouds are IMO the future of IT. Being able to “outsource” some resource pools from public cloud for lets say development, testing and low profile in general – is very attractive from enterprise perspective. Having this in mind I’d say: cloudbursting – is a fashion & cool feature whereby hybrid clouds are real need.

  2. Posted August 17, 2011 at 8:28 pm | Permalink

    Hi Sebastian,

    Thanks for your comment! The cloud “outsource” model definitely has it values. When used in this model, I think it’s better called cloud extension than cloud bursting. To me, cloud bursting is more about real time migrations of workloads, which are not quite necessary for normal use cases in the outsource model. Having said that, real-time is always better than otherwise when cost is not an issue. :-)

    Steve

  3. Posted August 19, 2011 at 5:00 pm | Permalink

    I would agree that while cloud bursting is good in theory, in practice it is complicated by bandwidth costs, latency between the clouds, and security issues (once your bits and bytes leave the “bursted” cloud into the “burstee”, they are in the Wild West of the public Internet). It is for this reason that a lot of customers we talk to come into a conversation believing that cloud-bursting is the solution to all their problems, but leave the discussion realizing that it brings as many issues as it solves.
    The hybrid cloud appears to be a more promising model, in that it allows for more flexibility in types of workloads, as well as simplifying certain requirements. For example, an application that must be PCI-compliant could house its database and storage components in a private cloud, while running web servers and other non-security-critical components in the public cloud. There are still issues with securing traffic that must cross the cloud boundaries, but many constraints can be removed by segregating application tiers between the infrastructures within the hybrid environment.
    – Brian Adler, PS Architect, RightScale

  4. Posted August 21, 2011 at 3:26 pm | Permalink

    Hi Brian, thanks a lot for your input and feedbacks from customers.

    Steve

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*

You may use these HTML tags and attributes <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

  • NEED HELP?


    My consulting helps clients with virtualization and cloud computing, including VMware infrastructure automation and orchestration, vSphere management APIs, and deep product integration with hypervisors. Current training offerings include vSphere APIs training, vCenter Orchestrator training, and etc. Should you, or someone you know, need these consulting services or training, please feel free to contact me: steve __AT__ doublecloud.org.

    Me: Steve Jin, VMware vExpert who authored the VMware VI and vSphere SDK by Prentice Hall, and created the de factor open source vSphere Java API while working at VMware engineering. Companies like Cisco, EMC, NetApp, HP, Dell, VMware, are among the users of the API and other tools I developed for their products, internal IT orchestration, and test automation.