Kathy Gibson reports from the Huawei Cloud Congress in Shanghai – System downtime is an expensive business: it costs businesses millions of dollars every year, to the extent that about 10% of R&D in the ICT industry is spent on disaster recovery.
Avneesh Saxena, group vice-president domain research: Asia-Pacific at IDC, explains that businesses today has new expectations from their ICT.
ICT’s role is now driven by business outcomes, he says. “Businesses now invest in ICT because they believe they can achieve certain outcomes, while ICT is expected to help the business come up with new products and services.”
In the new environment, Saxena points out that ICT systems must not only be more fluid and agile, but more resilient as well.
“Going forward, there will probably never be a moment when the business can live with outages. We have to make sure we are constantly running and constantly aligned,” he says.
In addition, data centre complexity is constantly increasing, driving a need for more and better automation to cope with massive amounts of data and management across the system.
The recent IDC Transformative Index found that more than 70% of IT organisations are less than 50% able to cut over to their DR site in realtime. More significantly, a Ponemon study puts the average cost of downtime is no placed at $7 900 per minute, up from $5 600 per minute in 2010.
“We believe the tolerance to downtime will go away in almost every industry,” Saxena says. “The user just expects IT to be always there, and the bandwidth to be always there. Expectations are very different to what they used to be.
One way that IT can address the increasing challenges associated with downtime is through cloud-based solutions, he adds. And, while IDC believes the industry is still some years away from a perfect solution, some vendors – notably Huawei – have made a pledge to openness and started rolling out solutions.
“Building resilient IT is as critical as power and communications for a business,” Saxena says.
Patrick Zhang, president: marketing and solutions at Huawei Technologies, points out that disaster recovery solutions generally run into three key challenges.
“Everyone builds a DR centre, but it is seldom used,” he says. “So it becomes a redundant asset and we need to find ways to make the backup data centre active.
“In addition, we don’t use the backup data centre very often, so when a disaster occurs, we need to be very sure of its operational ability and that the service is consistent. Because we don’t have visibility into the backup data centre,
we are afraid to cut over.
“The third challenge is in how to cut over. Sometimes it takes too long and is too complicated. And the longer the downtime the bigger the cost, especially for critical businesses that have zero tolerance for downtime, like financial services and railways.”
Huawei’s active-active disaster recovery centre addresses many of the challenges associated with resilient systems, Zhang says.
“For backup assets, the active-active DR centre provides a solution to computer resources at five levels, making them all active.” An example would be the virtualised storage centre which can write data in both the active and backup data centre simultaneously, regardless of the equipment vendor in either centre.
“The Huawei Active-active DR centre supports heterogeneous, multi-vendor data centres, recognising that the active data centre may have different systems from the backup centre,” Zhang says.
Visibility into the backup data centre can be solved with a virtualised systems that enables a single management view.
“Using the FusionSphere cloud operating system allows IT to virtualise both the active and backup data centres into one resource pool, so it can be managed as one data centre,” he says. “In addition, all applications are active-active as well, running in both the active and backup data centres. So if one data centre goes down, the applications and data continue to run in the other.”
The challenge of how to cut over to the backup data centre is addressed through the ManageOne management console. “Normally the DR centre is seen as a black box, but we need to know about its competence and performance, and check the data consistency on both sides. ManageOne gives the IT manager visible management into both active and backup centres, so when the business has to cut over he knows for sure that business operations are the same on both sides,” Zhang says.
Zhang officially launched the Huawei Active-Active DR centre today, pointing out that the company is building on a rich product portfolio that combines IT networking, communications, telecommunications and energy solutions.
Orchestrating the active and backup data centres requires low latency, and Huawei also announced its long-distance WDM which is able to connect data centres up to 300km apart with a latency of just 5ms.
Zhang adds that Huawei knows the Active-active DR centre solutions works – because the company has been using it in its own data centres,
“We have three main data centres and operate with node data centres in 150 countries,” he says. “This distributed data centre is provided by Huawei, so we are the guinea pig for our own new technologies.”