No-one likes to wait. Not customers, not staff, not business partners. So what can be done to improve whole-of-company IT performance?
One of the least known, yet lowest-hanging fruit is to interconnect company IT directly to hyperscale cloud providers via lightening fast, private links according to NEXTDC, Australia’s leading independent data centre operator.
Once the variability of public Internet connections are taken out of the mix, troubleshooting performance issues becomes radically easier.
The company says many Australian companies are yet to take up the opportunity to directly connect to hyperscale cloud with consistent, unwavering latency measured in milliseconds.
Once connected this way, cloud applications can perform as consistently as those hosted in corporate data centres, but without the scaling challenges and high infrastructure costs.
NEXTDC’s business is centred around building the critical infrastructure “where the cloud lives” and hosts more public cloud infrastructure in Australia than any other data centre network.
What this means is businesses with IT infrastructure colocated in their data centres can literally plug straight into public cloud infrastructure for vastly improved performance, reliability and cost efficiencies.
Australian companies are increasingly directly connecting in to hyperscale cloud providers, driven by a desire to improve customer experience to climb above global competition. A finely-tuned, consistently-performing hybrid cloud environment is one of the keys to unlocking that objective.
According to NEXTDC, making the changes outlined here can bring your IT stack closer to your cloud platforms, while also driving both upfront and long-term financial gains.
COVID-19: A catalyst for business collaboration
Digital transformation, or workplace digitisation is a work-in-progress for organisations, but the sudden emergence of COVID-19 shone a spotlight on the importance of a responsive and resilient IT stack to help companies adapt quickly to a suddenly different world.
Every CIO is now looking to their IT team to find creative new ways for the company’s workforce to collaborate efficiently across a distributed environment, under even tighter cost constraints than usual.
Fortunately, most businesses have made – or are part-way through implementing – a Hybrid IT stack, which allows them to retain control while benefiting from the infinite elasticity of cloud resources to scale performance and access new features without on-premises constraints.
This has enabled many companies to pivot to remote-working on a much grander scale than ever before – and it’s likely that many of these new work practices will remain long after COVID-19 is brought under control.
Hybrid IT isn’t a “set and forget” architecture – understanding the architectural options and making the right design decisions can make all the difference when it comes to collaboration, performance, and cost efficiencies.
The bottom line is that removing friction across the whole workforce by enabling a smoothly running IT machine is key to continuing to deliver improved customer experience, even amidst volatile market conditions.
You are in how many clouds now?
Anyone working in IT knows that the perception that a company is only connecting to a small and carefully controlled number of public clouds is simply not the case. When a thorough audit peels the lid off the sardine tin, it invariably shows the organisation is squeezing in many more than they thought.
NEXTDC’s Head of Product, Adam Gardner, explains, “If you look deeply at the IT process, many organisations use a wide range of cloud-based applications. This could include Office 365, Microsoft Teams, Salesforce, Slack, SAP, Dropbox, Zoom, AWS, Google Cloud and the list goes on.
“In fact, each of us uses a vast array of different cloud services. Each of those may only function as 5% of our daily processes, but if they’re down or not performing well, how much time is wasted across the workforce trying to troubleshoot or workaround the problem? Often these systems are situated within a workflow, so individually it is an annoyance, but cumulatively it can destroy productivity and output”.
Why speed matters
Putting a dollar figure on poor application performance can be elusive, however, tech giants have been able to quantify it. Greg Linden, now a Data Scientist at Microsoft, worked on performance testing at both Google and Amazon and can equate slowness to a real dollar value.
His results were incredible: each 100ms that Amazon.com was slowed down resulted in a measurable drop of 1% of sales in A/B testing (where a control group gets the usual experience and another group gets the modified site).
At Google, tests ordered by Marissa Meyer, VP of Search and User Experience, to artificially slow down Google search found that adding just half a second (500ms) of delay to Google search resulted in 20% less traffic.
The lesson, Meyer, said, was
Speed matters. People do not like to wait. Do not make them.
Industry titan Steve Jobs made a compelling back-of-the-envelope calculation in 1983 when he was developing the first Apple Mac. "The Macintosh boots too slowly. You've got to make it faster!" Jobs exhorted to the team. "You know, I've been thinking about it. How many people are going to be using the Macintosh? A million? No, more than that. In a few years, I bet five million people will be booting up their Macintoshes at least once a day."
"Well, let's say you can shave 10 seconds off of the boot time. Multiply that by five million users and that’s 50 million seconds, every single day. Over a year, that's probably dozens of lifetimes. So, if you make it boot 10 seconds faster, you've saved a dozen lives. That's really worth it, don't you think?"
The team did manage to shave more than 10 seconds off the boot speed of the Mac in the months ahead of launch, “saving lives”.
Unnecessary network latency in applications is the 2020 equivalent of the floppy disk load times the Apple team struggled with in 1983. While the impacts are measured in milliseconds not seconds, poor latency can quickly add up to meaningful impact.
Imagine 250 staff connecting to an application via the Internet. The software is making 1,000 network transactions per user each day. Connecting via the Internet instead of a direct connection may add 30ms per transaction or 30 seconds per day. In isolation, that sounds irrelevant, however over a year and across 100 staff, that’s over 50 hours of staff time – for that application alone – that could be freed. Now, imagine how many different applications all your staff are using across the business to get a picture of the scale of productivity loss that tiny inefficiencies can cause.
How jitter destroys application performance
While the size of a connection (bandwidth) and the time it takes for a packet to make a round trip from point A to B (latency) are generally well understood in the IT world, jitter is less understood.
Put simply, jitter is the variation each packet experiences in latency. This happens when packets are being routed across uncontrolled public Internet routes, where packets bounce from network to network until they finally reach the destination. Each packet doesn’t necessarily take the same route – it will go wherever internet protocols determine is the best route at that precise moment.
If the first packet takes 50ms, the second 80ms and the third 60ms, there’s a high likelihood the first and third packet will arrive before the second one does. When data is received out-of-order at the other end, it puts software under pressure which has a noticeable effect on application performance and quality.
NEXTDC Head of Network Operations, Sean Rinas, explains, “video conferencing and voice calls, for example, are heavily impacted by jitter. When you see the picture freezing or breaking up, or voices dropping out for a moment, that’s more often due to jitter, not lack of bandwidth.
“Software has trouble dealing with out-of-order packets. Strategies like buffering are needed to attempt to overcome that, to give the software time to reassemble the packets into the right order. Unfortunately, the very high-performance hardware needed for network buffering is expensive and has physical limitations.
Getting network jitter under control through a direct connection to a cloud provider through a data centre provider like NEXTDC, virtually eliminates network jitter altogether because your direct path is always the same.
“When your packets are being received in the same consistent and expected order, it removes the need for your software to deal with decoding out-of-order packets. In turn, your application performance drastically improves, as does your user experience”
In an environment where our virtual connections are our only form of collaboration and therefore crucial, this represents a close to perfect delivery of user experience.
Translating speed into productivity
Most businesses are deeply reliant on cloud applications and don’t realise how small inefficiencies in their architecture all add up, which reduces the productivity of their entire workforce.
As an organisation do you want to be paying your team members to sit and wait for things to happen, or do you want to be investing a small amount into better architectural choices to let your team get on with productive work that will make real, incremental gains for your business and your customers?
explains, “Humans can easily appreciate 200ms latency. When we’re talking to someone, we expect a seamless transmission and receipt of signals. Just a fifth of a second delay is jarring; it’s enough of a delay for the seamless flow of communication to be disjointed.”
The closer you can get, and the fewer hops you have to take to get to the services you use, the better the experience. The better the experience, the better the output from your organisation with less friction and fewer delays.
While delays like this are most obvious in a voice or video call, the same micro slowdowns affect productivity and performance of all corporate applications. Finance systems that take a second to refresh a page vs 100ms can measurably affect throughput across dozens or hundreds of people using that system.
Tapping into where the cloud lives
NEXTDC’s headline is that it is “where the cloud lives”. As abstract as this may sound, it is in fact a literal reflection of the fact that NEXTDC hosts more hyperscale public cloud services in its data centres than any other data centre in Australia.
That means when you colocate your physical IT footprint in a NEXTDC data centre, you are given immediate capability to directly connect to your cloud platforms and other ICT services your business consumes.
This not only enables the most direct connectivity path to your cloud and ICT services; it drastically improves the performance and reliability of your applications and workloads.
It’s the closest you can get to your critical digital platforms, which is effectively the same outcome as bringing your critical business platforms to your office location, but without any of the costs associated.
When your physical IT is colocated within the same facility as your cloud services, you’re only milliseconds, if not milliseconds, away from the likes of Microsoft, AWS, Google, Oracle, IBM and Alibaba Cloud.
NEXTDC has a huge, national, software-defined network called AXON which connects its partner ecosystem of more than 600 clouds, networks and ICT services partners together. It even extends into other data centres, where hyperscale cloud providers are hosted to ensure our customers are given complete choice and flexibility as to how they architect their Hybrid IT stack.
“There might be a scenario where a customer needs AXON connectivity to another major commercial or government data centre. We proactively invest to make our competitors locations and other public infrastructure accessible to our customers, which we commit to under the standard AXON performance SLA” reflects NEXTDC Head of Network Operations, Sean Rinas.
“Whilst some data centre providers choose to ring fence customer options, we genuinely try to remain flexible, and help our customers get to where they need to be with as little friction as possible, rather than sweating commercials. Although that may sound a bit cliché, we believe that’s what’s needed to build strong customer relationships long-term” agrees NEXTDC’s Head of Product, Adam Gardner.
Making a direct cloud connection
“Private fibre to the cloud” sounds horribly expensive – and it could be if it was truly a point-to-point connection to a data centre – but the truth is, due to the enormous capacity already available in existing fibre networks, it’s not necessary.
Most large organisations are already in buildings served by telco fibre and getting their internet and WAN connectivity over it.
Adding a direct connection to cloud infrastructure, such as a NEXTDC data centre, is simply a matter of getting the telco to configure a virtual tunnel within the fibre path. This can deliver practically the equivalent of point-to-point dedicated fibre, at a fraction of the cost.
Alternatively, customers can relocate their physical IT footprint in the data centre and from there a standard Cross Connect will get you directly connected to the cloud infrastructure your company uses. Performance-wise, it’s the closest thing to having AWS or Azure in your office building, but with none of the responsibility or cost associated with managing it.
If you’re in a NEXTDC data centre, you can connect to these hyperscale cloud providers via a single Cross Connect - $120 a month for example would give you up to 10GbE connectivity into that provider. But if you tried to buy 10GbE internet for your office, it could be well over $10,000 a month.
Taking the right turn toward cloud
If your company is connecting to any of the public clouds without a direct on-ramp to that cloud provider, you open the door to inconsistent connection experiences caused by the public Internet, rather than you maintaining control of performance.
NEXTDC has the largest number of direct-to-cloud on-ramps in Australia, allowing customers to connect directly into hyperscale cloud providers like Amazon, Microsoft, Google and near hyperscale providers like Alibaba Cloud, Oracle and IBM.
If you’re in Sydney or Melbourne, you can connect to a NEXTDC data centre and have all your traffic from there routed directly into the cloud providers with millisecond latency.
In other states, many of NEXTDC’s data centres also have on-ramps to the cloud providers. For example, Perth customers can connect to NEXTDC’s Perth P1 data centre and connect directly into Microsoft or AWS’ only Perth-based on-ramps.
This represents two key advantages for businesses: firstly, once you access the cloud on-ramp, your traffic is routed to the cloud provider via an optimal route on NEXTDC’s lightning fast national network, with minimal latency and jitter.
Secondly, the Service Level Agreement (SLA) with the cloud provider starts from the moment you access that cloud on-ramp. This directly translates into more guaranteed uptime for your cloud services, and less risk that comes with the unreliability of the public Internet.
Let cloud do the heavy lifting
The point of cloud is to let massive computing resources do the data crunching in huge data centres, away from your own server room. For pure cloud instances, that philosophy holds true.
But the prevalence of Hybrid Cloud in most companies means data is accumulating and being processed on-premises, and then overflow workload is being shuttled to the cloud.
This means that the link between the company’s own data centre and the cloud is a limiting factor, both in terms of responsiveness (latency) and speed (throughput of the link).
By colocating your physical infrastructure at a hyperscale data centre like NEXTDC, you have the benefit of all the data accumulation and workload happening in the most resilient and reliable facilities available, with extremely fast connectivity to the cloud services taking the overflow workload.
‘No strings attached’ connectivity
Leasing high performance carrier fibre services often comes with long contracts to help the telco amortise the infrastructure spend over time. Terms of 36 months are not uncommon.
NEXTDC built the “AXON” network, which is a software defined networking ecosystem, that provides dynamic Connectivity-as-a-Service.
Gardner explains, “If you have an AXON port, you can provision multiple virtual connections in just a couple of minutes.
“Let’s say you have a big data transfer event on the weekend – a migration of your company’s data from an Exchange server to Office 365 for example – you can spool up a 10GbE connection to Microsoft on the Friday and shut it down on Monday.”
This Connectivity-as-a-Service isn’t only available within core data centres in Sydney, points out.
“If you’re in Brisbane, where there are no hyperscale clouds hosted, and you need to transfer a large amount of data, you can log in to the AXON portal and turn on a dynamic intercapital connection within seconds for the time frame it’s needed.
In NEXTDC’s data centre network, there are over 70 carrier or carrier-like partners available. In a standard office building, there may only be one or two carriers available to service the building, Gardner explains.
The simple fact is customers have more choice and more flexible arrangements that will drive more performance and cost efficiency gains for your business.
Ensuring connectivity for your business without the headaches
One of the biggest advantages for businesses is NEXTDC’s 100% uptime SLA, which is unique in the Australian market.
But tradies putting backhoes through fibre paths outside the data centre – that’s a perennial problem that no-one has figured out how to prevent. As recently as May 2020, a major Internet and mobile network outage that blanketed Melbourne’s eastern suburbs was tracked to a construction crew. The unwitting contractors had cut through a Telstra fibre-optic cable path, causing “extensive” damage that repair crews had to work through the day, night and following morning to repair.
For businesses looking to safeguard connectivity, having two geographically diverse network paths is the workaround – if one goes down, the other should still be up. But it’s not as simple as it sounds.
“Telcos lease fibre from each other all the time rather than putting new fibre into the ground, and they aren’t usually willing to disclose their cable maps, so it can be impossible to determine that a path they’re proposing is truly independent from another path,” explains NEXTDC Head of Network Operations, Sean Rinas.
NEXTDC can solve that problem with its AXON network and Data Centre Interconnect (DCI) service, which provides customers with guaranteed physical diversity.
“The customer gets different interconnect rooms, different fibre in the ground, different internal cabling, different racks, all the way until we hand-off to the customer,” Sean explains.
“For customers who want two physically differentiated services, we make that headache go away.” he adds.
“The NEXTDC team literally stand there when fibre is being installed,” recollects Sean.
We have absolute confidence that diversity can be guaranteed, which is the level of detail we go into when it comes to standing by our commitment of 100% uptime for our customers
Making the connection
As you look to better align your Hybrid Cloud architecture to enable your business strategy, partners like NEXTDC can help support your objectives of building flexibility, versatility and resilience into your environment, allowing you to save money on the total economic impact of in-house infrastructure.
The money you save on unnecessary real estate, inefficient network links and resources will free you up to focus on what matters most, your business, and making the final push to complete digitisation.
Talk to NEXTDC to find out how you can optimise your Hybrid IT architecture for maximum performance, minimum friction, and improved customer experience.Contact NEXTDC