Cookie policy: This site uses cookies (small files stored on your computer) to simplify and improve your experience of this website. Cookies are small text files stored on the device you are using to access this website. For more information on how we use and manage cookies please take a look at our privacy and cookie policies. Some parts of the site may not work properly if you choose not to accept cookies.

sections

Cracking latency in the Cloud

  • Print
  • Share
  • Comment
  • Save

Latency has always been a challenge when developing applications and developers and networking engineers who work in the Cloud now have to understand a whole new set of latency delays. Here are some of the common issues and some advice on how they can be overcome.

The cost of latency

To understand latency, we need to understand the costs of it and how it impacts business.

There have been some studies that have examined overall website performance with respect to latency. For example, every drop of 20ms of network latency will result in a 7-15% decrease in page load times.

This study isn’t just an academic exercise: both Amazon and Google found big drops in sales and traffic when pages took longer to load. A half-second delay will cause a 20% drop in Google’s traffic, and a tenth of a second delay can cause a drop in 1% of Amazon’s sales. It’s not just individual sites that have an interest in speeding up the web.

While reducing latency is desirable, not every app requires the lowest latencies. Certainly, we have become more demanding of our internet performance as we distribute our applications throughout various cloud-based providers. Applications such as algorithmic or high-frequency trading, video streaming, more complex web and database services and 3D engineering modelling are in this category. But applications such as email, analytics and some kinds of document management aren’t as demanding.

Possible pitfalls

In the past, latency has had three different measures: roundtrip time (RTT), jitter and endpoint computational speed. Adding traceroutes as a tool, each of these is important to understanding the true effect of latency, and only after understanding each of these metrics can you get the full picture.

RTT measures the time it takes one packet to transit a network from source to destination and back again, or the time it takes for an initial server connection. This is useful in interactive applications, and also in examining app-to-app situations, such as measuring the way a web server and a database server interact and exchange data.

Jitter is a variation in packet transit delay caused by queuing; contention and serialization effects on the path through the network. This can have a large impact on interactive applications such as video or voice.

The speed of the computers at the core of the application is also crucial, as their configuration will determine how quickly they can process the data. While this seems simple, it can be difficult to calculate once we start using cloud-based computer servers.

Finally traceroute is the name of a popular command that examines the individual hops or network routers that a packet takes to go from one place to another. Each hop can also introduce more or less latency. The path with the fewest and quickest hops may or may not correspond to what we would commonly think of as geographically the shortest link. For example, the lowest latency and fastest path between a computer in Singapore and one in Sydney Australia might go through San Francisco.

Virtual complexities

Today’s modern data center isn’t just a bunch of rack-mounted servers but a complex web of hypervisors running dozens of virtual machines. This introduces yet another layer of complexity, since the virtualized network infrastructure can introduce its own series of packet delays before any data even leaves the rack itself.

But we also have to add to this the delays in running virtualized desktops, if that is relevant to our situation. Many corporations have begun deploying these situations and that introduces yet another source of latency. If not designed properly, you can experience  significant delays with just logging into the network, let alone running your applications on these desktops.

Relying on the Internet for application connectivity in the cloud introduces a degree of variability and uncertainty around bandwidth, speed and latency. This can be unacceptable to many large and medium-sized enterprises, which are increasingly putting the emphasis on end-to-end quality of service management. Using dedicated connectivity to cloud providers overcomes this and hooking up via carrier neutral data centers and connectivity providers can also have benefits in terms of costs, monitoring, troubleshooting and support.

Latency in itself does not have to be an issue; it’s the unpredictability of latency that really causes the problems. Although there are several complicating factors, careful planning can ensure that these pitfalls are avoided and applications run smoothly in the Cloud.

 

Related images

  • network file shot

Have your say

Please view our terms and conditions before submitting your comment.

required
required
required
required
required
  • Print
  • Share
  • Comment
  • Save

Webinars

  • Do Industry Standards Hold Back Data Centre Innovation?

    Thu, 11 Jun 2015 14:00:00

    Upgrading legacy data centres to handle ever-increasing social media, mobile, big data and Cloud workloads requires significant investment. Yet over 70% of managers are being asked to deliver future-ready infrastructure with reduced budgets. But what if you could square the circle: optimise your centre’s design beyond industry standards by incorporating the latest innovations, while achieving a significant increase in efficiency and still maintaining the required availability?

  • The CFD Myth – Why There Are No Real-Time Computational Fluid Dynamics?

    Wed, 20 May 2015 14:00:00

    The rise of processing power and steady development of supercomputers have allowed Computational Fluid Dynamics (CFD) to grow out of all recognition. But how has this affected the Data Center market – particularly in respect to cooling systems? The ideal DCIM system offers CFD capability as part of its core solution (rather than as an external application), fed by real-time monitoring information to allow for continuous improvements and validation of your cooling strategy and air handling choices. Join DCIM expert Philippe Heim and leading heat transfer authority Remi Duquette for this free webinar, as they discuss: •Benefits of a single data model for asset management •Challenges of real-time monitoring •Some of the issues in CFD simulation, and possible solutions •How CFD can have a direct, positive impact on your bottom line Note: All attendees will have access to a free copy of the latest Siemens White Paper: "Using CFD for Optimal Thermal Management and Cooling Design in Data Centers".

  • Prioritising public sector data centre energy efficiency: approach and impacts

    Wed, 20 May 2015 11:30:00

    The University of St Andrews was founded in 1413 and is in the top 100 Universities in the world and is one of the leading research universities in the UK.

  • A pPUE approaching 1- Fact or Fiction?

    Tue, 5 May 2015 14:00:00

    Rittal’s presentation focuses on the biggest challenge facing data centre infrastructures: efficient cooling. The presentation outlines the latest technology for rack, row, and room cooling. The focus is on room cooling with rear door heat exchangers (RHx)

  • APAC - “I Heard It Through the Grapevine” – Managing Data Center Risk

    Wed, 29 Apr 2015 05:00:00

    Join this webinar to understand how to minimize the risk to your organization and learn more about Anixter’s unique approach.

More link