News

The driving power behind the cloud

Microscope contributor


As many companies move into the cloud there is a growing need for tools that can monitor, analyse, optimise and secure the supporting infrastructure.

There are a number of benefits to be gained from cloud computing, be it an internal cloud within your company, or external through a third party with data accessed over the web. The growing complexity of dealing with “fat clients”, maintaining multiple software systems and complicated IT infrastructure has convinced many that the centralised, thin client cloud model is the right path to take.

But the price of cloud computing is control. When adopting a cloud computing model, control of critical applications and systems is passed to the cloud service provider. Companies must accept this. Given the interest in cloud computing, it would seem that this mental hurdle has been accepted.
But if this acceptance and underlying confidence is to be maintained, the cloud computing infrastructure needs to ensure that critical applications and systems are available at all times.

Therein lies the rub, because cloud computing, like any other technology, is not infallible, as highlighted by recent hiccups in Google’s Gmail service. Companies have to feel confident that their applications and data will be available and secure at all times. As a result, there is increasing interest in tools for monitoring, analysing, optimising and securing the application delivery infrastructure.

While there is a great deal of focus on cloud computing services and the possibilities they present, it is also important to consider the delivery infrastructure and make sure it provides the availability and performance required.

Guaranteed delivery

Availability is the number one issue for cloud computing. Companies and customers want to know that they can access their information at all times and that it is securely stored. The challenge is to meet that expectation and know how to mitigate potential problems.

The worst scenario is that you cannot access your data. After that, it is a question of time and how long you are without access.

Cloud service providers can ensure that their data farms have redundant systems with regular back-ups and users can ensure that they have a reliable internet connection, but the key connection between the two is often outside both the user’s and the service provider’s control. Cloud service providers should monitor their connections to ensure optimal use and adequate protection. They should also demand high availability from their network connectivity provider, with detailed reports on performance.

This demands investment in dedicated network appliances capable of monitoring and analysing high-speed connections in real-time. Not everyone is familiar with these devices or understands what they can provide, but for those responsible for managing networks these dedicated network appliances are becoming critical tools for ensuring the availability of the network and the services supported by it.

Speed and efficiency

Another piece of the puzzle is performance and the need to perform in real time. As with availability, if links are not working and customers have to resend information, they will not be happy. The success of cloud computing demands that users experience the services as if they were on the desktop and are not reminded of the physical distance to the service by experiencing delays.

The nature of cloud computing as a centralised model with multiple users accessing the same service at the same location means there will be critical points in the network where a lot of people are trying to get a lot of data through very few connections, resulting in potential traffic jams.

This, in turn, means larger bandwidth connections running at higher speeds with higher line utilisation. Network monitoring and analysis tools are required to ensure that there are no performance issues, but these tools must also be capable of handling the vast amount of real-time data that will flow through these connections.

There is an increasing trend towards using dedicated network appliances that specialise in these tasks. As we move to cloud infrastructures there are requirements for these network appliances to handle vast amounts of data in real time.

For example, if an application needs to monitor and analyse all data traffic on a 10Gbps port, it needs to be capable of handling up to 15 million packets per second – or a packet every 67ns. These network appliances must be able to see all the services, analyse them and take action quickly. The focus is on pre-empting problems rather than reacting to them, as once the problem occurs, it is already too late.

The key performance criteria for network appliances are speed and efficiency. The appliances must have the horsepower to process high-speed data in real time while not using a lot of server processing power, and do this cost effectively. Fortunately, there are solutions that meet all of these criteria.

Traditionally, network appliances have been built using proprietary hardware and architectures. This has provided high performance, but at a higher cost. But advances in standard server architectures and processing power allow standard servers to be used to build high-performance network appliances that are very cost-effective. To ensure high performance, you simply need to add a network adapter that can handle vast amounts of data in real time, without losing packets or introducing excessive delay, while using very little processing power.

The types of devices that have been based on such a model are network performance management appliances, network forensic and diagnostic tools, network test systems, network security solutions, and network latency measurement systems.

These products are important in assuring the availability of the cloud computing infrastructure and are capable of meeting present and future cloud demands.

Maintaining services

When moving to the cloud, focus tends to be on the service and its provider, with little thought given to the infrastructure in between. It is assumed that it will always be there and that access to the cloud service is a given.

But this is not automatic. Investment in dedicated network appliances for monitoring, analysing, optimising and securing the network are essential to maintain the availability of cloud services.

While cloud computing is placing higher demands on performance, network appliance vendors are equal to the challenge and are capable of providing high-performance, real-time solutions that are based on cost-effective standard server platforms.

By using these network appliances, firms can ensure that the journey into the cloud is smooth and uneventful.


Join the conversation Comment

Share
Comments

    Results

    Contribute to the conversation

    All fields are required. Comments will appear at the bottom of the article.