Warakorn - Fotolia

Get started Bring yourself up to speed with our introductory content.

Identifying the top 10 network-cloud integration challenges

IT managers face a variety of network-cloud integration challenges. But there are steps they can take to keep those challenges at a minimum.

Your application performance is only as good as your network. New applications are placing increasing demand on corporate networks. Mobile devices, big data and customers' transactions are increasing network-cloud integration challenges and driving the need for purpose-built networks that can scale as business demands require. Here's what to look for in a service provider to ensure your network capacity meets your needs:

  • Ethernet everywhere: The ability of a network operator to maintain a consistent level of service around the globe is key for enterprises. Ideally, your network service provider should be able to deliver the same "Ethernet everywhere" experience from all regions. Some operators have a strong presence in Asia, but not in Europe; others are strong in the United States but not in Latin America. The globalization of business requires service providers to standardize their service delivery across all markets to meet the needs of multinational corporations.
    As the saying goes, your cloud is only as good as your network.
  • Real-time network analytics: It is also important for IT professionals to understand and analyze application performance issues as they relate to the underlying network. Many of the most innovative network services provide a dashboard of network statistics and the ability to perform real-time analytics of network performance. Having visibility into application bandwidth performance requirements will enable IT departments to buy just the bandwidth they require for particular applications and specific periods of time. Many businesses have seasonal peaks and require the flexibility to adjust their IT resources to meet changing business demands.
  • Bandwidth on-demand capabilities: Many large network operators lack the ability to deliver bandwidth on demand to customers. When implementing a service-level agreement, it's essential for businesses to require the network service provider to enable both planned bandwidth increases, such as when testing a new application, or unplanned events such as natural disasters.
  • Last-mile diversity: The so-called last mile remains a complex issue for large service providers and carriers. The last mile to the customer's premise can pose a serious bottleneck if not managed properly. Look for suppliers that have control over this portion of the network so that they are able to deliver consistently high quality-of-service (QoS) levels from LAN to LAN. Large telecommunication service providers make their profits based on the volume of service they sell. It is always in their best interest to sell customers as much bandwidth as possible, even if the customer doesn't require a high level of bandwidth at all times.
  • Cloud marketplace bridging: With the growing adoption of cloud infrastructure services such as Amazon Web Services and IBM SoftLayer, the need to integrate the network with cloud compute services is essential. A company's cloud strategy should be based on business objectives. But we often observe that companies lose the economic benefits of the cloud if the corporate network lacks flexibility. The network must be able to meet the bandwidth and configuration demands of new cloud services and applications. As the saying goes, your cloud is only as good as your network.
  • Mobility and security: Bring your own device (BYOD) is generating a lot of networking challenges and security issues for IT departments. BYOD has become a widespread practice by employees. It's a good idea to sandbox traffic coming from smartphones, tablets and laptops to prevent these roaming devices from bringing malware and other harmful software into the corporate computing environment. The use of network packet inspection and network behavioral analysis can also help IT departments spot potential security breaches via mobile devices.
  • Global network operations centers: Administrators and engineers appreciate the ability to speak to a network expert at the service provider’s network operating center when problems occur. Yet many large operators outsource the first line of customer service to offshore call centers with support personnel who do little more than fill out trouble tickets.Some of the top network service providers offer proactive monitoring and operate as an extension of your IT team. These network experts can identify and remedy network issues before they become critical and impede the normal flow of business activity.
  • Service-level agreement for in-sequence packet delivery: Service-level agreements (SLAs) governing in-sequence packet delivery and jitter are major metrics that IT departments should consider when moving to a new operator. A company's SLA with a network operator should stipulate 100% in-sequence packet delivery and less than 1 millisecond jitter rates to ensure a high level of network performance to meet all application processing requirements.
  • Customizable solutions: Many large operators offer so-called catalog solutions that do not necessarily match the application and performance requirements of customers. One size clearly does not fit all. Look for an innovative operator that will craft a network for a customer that meets the businesses' unique needs and changing requirements. Make sure the provider does not need to take down the network or sell new services to enable it to adapt to the company's evolving needs.
  • Single- or multistrategy provider: Opting for a multiprovider strategy or single provider is a question IT professionals face when deciding to renew a contract or to switch to a multiprotocol label switching network. There is no right or wrong answer, but the approach should be as transparent as possible to avoid technical problems and procedural issues. The IT department should opt for the most transparent option that lets network managers identify and troubleshoot problems quickly.

About the author:
Ray Watson is vice president of global technology at Masergy Inc.

Next Steps

What to expect from cloud computing in 2015

Network infrastructure an Achilles' heel?

Lease or own your cloud?

This was last published in February 2015

Dig Deeper on Network Infrastructure

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

What cloud-network integration challenges has your organization faced?
One of the biggest challenges we ran into was with customizable solutions. We implemented a custom VPC with one vendor who later standardized and published their own VPC solution.  One of the cloud-based tools that we brought on was written specifically to work with the now default VPC, and did not work with a custom VPC such as we had implemented. We were able to overcome the issues, but only after many false starts, by working extensively with the cloud vendor and tool vendor to cobble together a working solution.
Our challenges have been with lack of network visibility at the application level,  and what tools should we use?

This topic is dead on,  my perspective is that "everything Cloud" rides the network,  either public network (Internet) or private network.

Application awareness at the network layer will be critical if this SDN thing is going to work.  Look what VMARE has done with virtualization of servers!  which btw rides the network too!

Networks that are application aware are the future and the ability to SDN like tune is key.
Great article...We have been doing business with a provider that recently entered the bandwidth on demand category only to find out that when we make changes within their portal, there is still a manual process behind the scenes.  This almost seems to defeat the purpose.  

We are now looking at providers leveraging true Software Defined Networking and Network Function Virtualization for our future network.  While there are many companies within the telecommunications world talking about these topics, very few are actually doing them.  We would like to ensure that our cloud strategy is supported by a robust network architecture while ensuring a true outcome based design.  
For us, much of it is coming down to the level of use and active management of the nodes in the cloud. Managing the machines and not having orphaned devices that rack up charges if left unattended can give me sleepless nights at times. Fortunately, we do a regular review, but it would be cool if we didn't have to always do that manual step.
Great article that is spot on.  Working with providers across the globe we have very little visibility especially since everything is moving to the cloud and no way to help end user experience.  Bandwidth on demand has been promised for years but it seems all the large carriers are trying to piecemeal a network together and this is still a long way to reality. Cant wait for Software Defined Networking to become a reality seems everyone is talking a good game but very little of them can implement anything close at this point especially globally.
The biggest challenge in cloud-network integration is, like all things involving the cloud, security. It is tricky to find an even balance between accessibility and security, especially with the risks that the cloud can pose. And always online open data source is almost certain to draw in attacks, and ensuring that they all fail is quite a job.
In my opinion, real-time analytics provides enterprises with continuous statistics, especially regarding bandwidth, while mobility and security measures scrutinize traffic from PC’s, tablets, and smartphones.