At times it feels like the concept of cloud computing is as ephemeral as a towering cumulus cloud on a summer day. It passes by, changes shape and looks different to every person who views it. But like those clouds in the sky, there’s a lot of complexity out there, especially as it concerns how data is managed in a multi-tenant environment.

The issue keeps coming up. First off, there are no existing standards for moving data in the cloud. Third party vendors like CloudSwitch are providing ways to “drag and drop,” data from on-premise to a cloud environments. But there is just no standardized way to move information between cloud networks.

But there are some ways to manage the complexity. Informatica’s David Linthicum wrote a blog post yesterday on the topic. We agree with Linthicum to some extent but his declarations don’t entirely reflect the current reality.

Linthicum outlines some of the inherent issues that come with data integration in the cloud. The biggest issue comes down to the cloud computing providers. They do not provide syncing back to the customer’s on-premise environment. Linthicum points out that Salesforce.com is the exception to the rule in this regard.

So, according to Linthicum, what should the customer consider?

• Backup: This is the first and foremost issue for Linthicum. You need to have an on-premise backup in case of cloud computing outages,cloud service providers going out of business or to prevent the headaches that can happen when the new owner of your cloud computing provider decides that it’s really not interested in that business at all.
• Data Movement: It’s the data integration that drives processes between systems that may be in the cloud or on-premise. How do you manage all this information that may be dispersed over geographically distant platforms? It’s a different world than dealing with systems that exist in one data center.

We agree to a point but there are lots of various options to these issues. First off, it’s often a matter of what cloud computing platform you choose. The large service providers offer the safest bet. They are vested in the future of cloud computing and are developing ecosystems that consist of third-party partners. A number of these partners specialize in cloud-based application deployments.

And cloud-based security is making advances that make data loss as rare as if the data is on-premise.

Linthicum does offer some sound advice that may seem more like common sense but it is worth noting:

“First, consider the overall requirements of the business. Sounds obvious, but many who deploy cloud computing systems do not have a complete understanding of the overall business requirements.Second, focus on the holistic architecture, on-premise and cloud-delivered, including how they will and should exchange data to support the core business.

Finally, select the right data integration technology for the job, and do so only after taking everything into account. You’ll find that there are both on-premise and on-demand options, and in many instances you may have to mix and match solutions.”

There are a lot of options out there. We do not necessarily agree that the hybrid approach is the best way to go. There is no one formula. Standardizing on the cloud has its advantages.

But, still, the over riding issue is really how data passes between systems. Without standards for managing data transfer, the issues Linthicum points out will never go away.

Related posts:

1. Cloud Computing Is Disrupting Microsoft Office 2010 and its Profit Margins Microsoft President Stephen Elop says the cloud has created a…

Related posts brought to you by Yet Another Related Posts Plugin.

View full post on Web Technology Blog, Development and Social Media

To keep up with the growth of cloud computing and virtualization, networks keep evolving. But unlike Twitter’s Trending Topics, IT budgets don’t scale up. In fact one of the major initiatives in many IT shops is creatively reduce their own expense.

To get to a scalable cloud infrastructure where costs are contained, it sounds like the network industry is going to see a time where a “Linux” arrives on the scene. An open source alternative to building networks may disrupt the networking landscape and give network admins an open source network operating system.

Virtualization: It’s in the Network Too

Distributing workload across machines, storage, and environments has required networks to be smarter than ever. Now, the network needs to be intelligent enough to not only route traffic both a bridge and a toll-gate, but to also provision and de-provision all aspects of the environment at a moments notice.

Providers like Rackspace are in the business of using the network to optimize the performance of the entire data center. To be effective in keeping up with dynamic system provisioning, technical teams need access to all tiers of the computing environment to reduce operations overhead.

Hosting providers, such as Amazon Web Services and Rackspace innovate new integration patterns – including ones in the core of the network – to get their job done. Network operating systems that are open, like Extreme XOS enable large scale hosting providers to look deeper into networking gear and start to tune it themselves. And enterprises may follow this trend.

Servers Don’t Sleep at Night, but Applications and Admins Do

For a long time, networks have been used to detect the peers and devices. Many of us use the nearly ubiquitous DHCP (Dynamic Host Configuration Protocol), which is the the thing that automatically assigns IP addresses to a PC when plugging into the network.

In an analogy, there is a need for a “super DHCP” is needed that can keep up with the highly virtualized cloud infrastructure per virtual instance. To do this, engineers look deeper to find efficiencies in how the network talks to the hardware and software for the virtual machines.

A good example of benefit for this is where a resource has peak loads during the day. Due to natural usage, the applications compute power is not utilized during the night. Using monitoring and provisioning tools, the network can de-provision the extra hardware and offer it to another service. This “freeing up” allocation saves power and money.

This is a simple example of where virtual data center solutions are being innovated in the industry to figure out how to further timeshare the computing resources. The network has the ability to help manage the scale down to the moment is enabled by it’s reach to everything over IP (Internet Protocol).

The Open Network Wins, Developers Rule

Extreme Networks is betting IT leaders that have become very familiar with Linux and open source Hypervisors like XEN want to tweak the network. For the data center manager that wants to go into the core network engines innovate, there comes a need for APIs, SDKS, and open access libraries.

Extreme’s openness is in the form of web services, many offered that are offered as XML or CLI scripting that allow integrate tools into the core of the network via XML, and configure edge ports for security and VOIP access as dynamic provisioning.

The company offers a code workbench of its own to download widgets to plug into the network. Designed for the open source developer, it shares the familiar pattern that presides in open source community for application frameworks and operating systems code sharing.

Shown in the diagram, Extreme’s network offers real-time provisioning of code widgets in the network.

Play Nice: the Networks Worst Enemy May be Success

Will the network evolve to see an open source player that drives change in pricing and value?

In the rush to enable new efficiencies we wonder if this is an Apple A-HA moment in the making. The question seems to be can the giants in the space balance the fine line of better end-to-end experience of managing the environment and whether vendors do it best. If we follow the Apple example of industry success, and end-to-end play for the network may be in the cards.

Last month, Juniper announced it has created a new business group and commitment to an Junos ecosystem.

Juniper has made a big move towards open source innovation in it’s recent re-branding and at least to one analyst, John Furrier from Silicon Angle, seems to be suggesting that Juniper Judo’ing Cisco, like Google did with Open and Microsoft. That probably doesn’t feel the least bit nice to the market leader, especially when Cisco is priming it’s engines for changing the Internet forever.

Cisco announced opening IOS in 2007 in an effort towards compartmentalizing IOS as part of its overall movement into a more software based organization. With the complex series of network enhancements and feature sets, it will be interesting to see how Cisco views “open” vs. “customizable” and where the control lives for network management and up-time.

When visiting the Cisco IOS website today, we see the standard license and no clear mention of open source licensing. Cisco strikes the balance between open and controlled in it’s a approach to defining what an open network is and where networks will be encapsulated as services.

We wonder if Cisco deliver the capabilities to pull more traffic into it’s end-to-end range, while open networking APIs rise as part of the network service stack. With this market, it’s likely both. At very least, open networking has a role in determining the fate of the network and where territories are being defined.

The Cloud is a Network of Services

The cloud is defining a world where service orientation rules – both the

software and physical layers. And, it is breaking the rules of workload distribution, where network topologies are changing. The requirements of connecting the layer 2 and layer 3 networks, as well as IT leaders that are building solutions for mass scaling (enterprises or service providers) are evolving and being driven by an ability to be efficient at the workload level.

Extreme Networks Technical Brief, Dynamic Network Virtualization Overview, explains the value of plug and play network components in today’s topology.

“By leveraging Extreme Networks ExtremeXOS, a modular, edge-to-core operating system, and our extensibility frame-work including Universal Port Scripting and an XML interface, Extreme Networks is able to tightly integrate the switching network with the virtualization environment to create a virtualization-aware network fabric that automates the network-level virtualization required in next generation data center and cloud computing environments. This unique functionality enables Extreme Networks to provide seamless support of virtualization capabilities across the various hypervisor platforms, including Citrix ZEN, Microsoft and VMware. The highly integrated solution allows the Extreme Networks solutions to trigger responses to virtualization moves as they happen in the network by virtue of a tightly integrated XML-based network management framework.”

Extreme, and now Juniper, are moving in the direction of offering IT administrators control points in networks and protocols to optimize it opens the market.

It looks promising to give administrators vendor leverage in buying services without vendor lock, or waiting for feature releases from the vendor. And, it mirrors the open-source movement in bringing communities together to solve problems and build compatible services.

Open APIs may define the cloud’s network of the future for large hosting providers. We wonder if for the enterprise.

Related posts:

1. Do Open Protocols Bring Storage Costs Down? The move to virturalization leaves stone is being left unturned….
2. Gladinet Cloud Desktop Gladinet Inc. of the US is one of the pioneers…
3. IBM Launches Cloud Platform For Software Developers IBM announced today a set of cloud services that will…

Related posts brought to you by Yet Another Related Posts Plugin.

View full post on Web Technology Blog, Development and Social Media