Network and Application Performance in Cloud

April 24, 2013
By guest blogger - Matt Smith

The word “network” is thrown around by some people without proper regard for what’s involved when cloud computing/technology is brought into the picture. As puts it “As enterprises construct more complicated cloud computing environments, networking also becomes more complicated. The challenges of traditional networking only increase with ‘private’ cloud, ‘public’ cloud and ‘hybrid’ cloud deployment.“

In other words, a whole new set of challenges come with cloud-environment networking. Some of those hurdles/difficulties include:

  • A lack of network capacity (as in network latency) can downplay/negatively affect the scalability and flexibility of cloud technology.
  • If networking capacity between the LAN and the cloud is over-stretched, services and applications are impaired.
  • Provider-to-provider networking may present special problems, as usage patterns and Internet connections are more easily monitored through a network plan.
  • The utilization of multiple cloud providers in order to reduce risk may negatively affect networks (e.g., impair traffic).
  • In attempting to move/migrate VMs and data, network re-configuration may become necessary; in the process, you may encounter problems with domain name services (DNS) and internal configurations (e.g., database clients and load balancers dilemmas).
  • Network latency and bottlenecks may be difficult to skirt/avoid without an appropriate monitoring plan.
  • How much of a Priority is Network/Application Performance Amelioration?
    Simply put, any enterprise that does not keep up with the latest, most efficient technology is going to be left behind. This means that every department of an organization has to be subjected to a painstakingly-well-diversified technological make-over. A good example is bring your own device (BYOD) movement leading to ubiquitous mobile electronic devices.

    While these devices bring much risk for the organization, they are also motivating levels of productivity and employee-employer/employee-the-world connectivity never before seen. This helps to illustrate why enterprises need to maximize network/application performance. Such amelioration, however, is possible only if specific best practices are proactively pursued. Some such practices include:

    1. Regarding mobile devices, institute/set up gateways, firewalls, anti-malware/viruses, data-removal strategies, and strong passwords.
    2. Look to replace old data centers with difficult to optimize servers, as well as any out-of-date LAN technologies unable to integrate server networking to storage networking.
    3. Use a combination of virtual/physical appliances for end-user-experience monitoring while outside the cloud.
    4. Continuously verify service level commitment by service provider(s).
    5. Generate/utilize application-dependency maps, as/when needed.
    6. When possible, opt for browser-based, agent-less end-user-experience monitoring straddling different cloud models.
    7. Measure network performance by way of any network path in order to pinpoint whether a problem should be attributed to CSP or your environment.
    8. Conduct end-to-end transaction analyses in order to ascertain ASAP the source of performance problems.
    9. Quickly identify/manage performance bottlenecks in database tiers, application, web, etc.
    10. Ascertain if application performance is being negatively affected by the cloud, the network used to access such, or by your environment.
    11. Set up “real-time” alerts if/when throughput or network latency is negatively affecting performance.
    12. Cost-effectively manage bandwidth requirements in migrating applications to the cloud.
    13. Ascertain what effect cloud migration will have on application performance.
    14. Develop/utilize workable application dependency maps ensuring the correct component migration into the cloud.
    15. Quickly identify/address network misconfigurations.

    What Other Factors Affect Network and Application Performance?
    For organizations that are just getting into the swing of things, technologically speaking, the first thing is to make sure that the move into the cloud is smoothly and smartly made. Configuring networks and applications is the next big hurdle. In addition to the best practices listed above, the following components/factors can play a key role in establishing excellent cloud/network/application relationships.

    Service Level Agreements
    SLAs spell out the parameters of what will hopefully be a long-term, mutually-profitable relationships between cloud service providers (CSPs) and your organization. As such, they need to be well-thought-out, comprehensive and prepared with the input of all stakeholders (lawyers, IT administrators, CSPs, etc).

    Wide Area Networks (WANs)
    WANs have opened many doors in terms of what’s been possible even before the cloud has been opened to the public. Their importance cannot be overstated in terms of their being a bridge to the cloud. As such, you need to appreciate their usefulness.

    Network and application performance optimization will no doubt become one of the fundamental goals for proactive organizations in the future. Cloud technology may still need refining, perfecting, but, like the Internet, it is here to stay. More importantly, it will open new avenues of opportunity, discovery and productivity that have, till now, only been imagined.

    Author Bio
    Matt Smith is a Dell employee who writes to help raise awareness on the topic of Cloud Computing and other network management subjects.

    Tags: , , ,

    2 Responses to “ Network and Application Performance in Cloud ”

    1. William V. on April 24, 2013 at 4:22 pm

      Great post Matt!

      There is no doubt a ton of different moving parts to consider when deploying applications on the cloud. Most of the problems that you have pointed out can be resolved by dynamic site acceleration (DSA) solutions provided by content delivery networks. A fully integrated DSA solution will address performance issues, latency, SLA needs, monitoring etc… It is also much faster and cost efficient to deploy.

      What is your opinion on DSAs?

    2. Susan Bilder on May 8, 2013 at 2:33 pm

      The BYOD movement is becoming more and more popular. It allows employees more flexibility but it can cause headaches for IT. When the network is accessible from a mobile device, it needs to be accessible at all times since mobile access means that people are doing work at all hours.

    Leave a Reply