Didier Marcon, bullion Architect, Bull Innovative Products
Didier Marcon’s career has encompassed many roles: as a hardware designer for GCOS mainframes, CAD software designer, AIX® validation manager for Escala™ servers, x86 Itanium architecture project manager and architect for Bull’s latest generation enterprise server: bullion™.
By the end of 2012, over 50% of applications running on x86 platforms will be virtualized. This figure illustrates the massive interest that businesses are showing in a technology which guarantees more flexible and lightweight infrastructure, and is indispensable in a smart switch-over to Cloud computing. But if you look at the number of applications there are – a figure that is set to expand four or five times by 2015 according to Gartner – and the fact that currently only 20% of mission-critical applications have so far been virtualized, there is a long way to go.
Indeed, it is gradually becoming clear that current virtualization technologies, which depend on large numbers of blade servers, only partially meet the demand for consolidation and virtualization on a massive scale, as well as the high availability needs of critical applications and power-hungry applications such as ERP implementations and large databases.
Even worse, to repeat a famous observation: “Blade servers are not the solution to these problems, they are the problem.”
A new approach to virtualization is becoming essential
And that is what Bull’s engineers have been working on for over five years now: to design an infrastructure that effectively reconciles complexity, low energy consumption, high availability, scalability and extraordinary performance. And above all, “without piling on more problems”. To bring simplicity back into the heart of IT infrastructures.
Which is at the root of a new and comprehensive approach to virtualization and the Cloud, combining a dedicated consultancy offering – Advisory Services – and a totally new kind of server, bullion: the virtualization virtuoso.
Cutting the complexity of virtualized infrastructures and simplifying their implementation
Rather than masking the complexity with another layer of infrastructure management tools, Bull has taken care of the issue at source: bullion enables customers to optimize resource utilization by actually using only the resources strictly required to meet the Quality of Service (QoS) targets defined in Service Level Agreements (SLAs). With bullion, IT architectures have become simple again, with no need to add more networking or processors when applications actually only require more memory, for example.
Avoiding the kind of sprawling of small servers that characterizes so-called ‘scale out’ architectures is a decisive factor in ensuring Quality of Service, normally only found with proprietary servers. Our customers have recorded significant improvements in Total Cost of Ownership (TCO) and very rapid Return on Investment (ROI). For example, one customer achieved a reduction of over 48% and ROI in under six months, by virtualizing their SAP infrastructures on bullion servers.
In a ‘scale up’ type approach, where each server has a high capacity to increase its processing power, bullion has been breaking records with its overall capacity of up to 16 Intel® Xeon® processors, with pricing proportional to performance. So customers can grow their processing power in line with their requirements, on the ‘pay-as-you-grow’ principle. Rather than buying a chassis half filled with blades, with resource utilization of less than 15%, bullion grows in line with customers’ needs. For example, organizations can start with a small cluster of only three bullion servers, each equipped with four processors. When their system management tools show a utilization rate above a certain threshold, additional modular resources can be added to the initial three servers: each can grow from four Intel Xeon processors to eight or even 16, doubling or quadrupling performance without increasing the number of nodes in the cluster or the resulting administration costs. With its unrivalled ‘scale up’ capabilities, no matter what target performance is required, with bullion the IT infrastructure remains simple and easy to manage.
bullion delivers twice the level of scale-up performance compared to any other competitor, with its 16 processors, equating to 160 physical cores inside a single server, using a single VMware hypervisor. The unique scalability – from a minimal, 2-processor configuration to a 16-processor server – is achieved by using the Bull Coherent Switch (BCS), an ASIC chip designed by Bull to feature a memory caching system for the interconnect network between the processors, memory and I/O.
With its BCS architecture, Bull meets customers’ demands virtualize their business-critical applications and databases. Scalability within a virtual machine (VM) is now guaranteed with the use of VMware vSphere5, scaling up to 32 vCPUs and 1TB of RAM per VM. These monster virtual machines have proved to be much more efficient than a larger number of smaller VMs when running ERP packages and databases. In some cases, ‘monster VMs’ of up to 32 vCPUs not only enable customers to concentrate their business-critical applications onto a smaller infrastructure, but also to actually consolidate applications themselves.
For example, a customer that decided to migrate its Java Application Servers to bullion was able to consolidate 200 partitions each running a Java Application Server, into just 70 VMs with several larger Java Application Servers. To provide the headroom and scalability needed by these kinds of applications such as databases, the bullion architecture supports an appropriately balanced system with more processors, memory and more I/O than any other x86 system, providing the cornerstone to consolidate business critical applications within the most efficient IT infrastructure.
Superior reliability: memory management under the spotlight
The complexity of IT infrastructures and their administration increases with the number of objects to manage, especially with so-called ‘scale out’ infrastructures. This is increasing with new ways of consuming IT. The approach taken by bullion is to tackle the issue of complexity at its source. That means offering an infrastructure that really only consists of the resources required to meet SLA levels. The key is to limit the inherent complexity of the infrastructure.
But the reliability of bullion is also based on an uncompromising approach to RAS (Reliability, Availability and Serviceability) components and technologies. Such technologies are not usually available on standard x86 servers, and they significantly boost server availability rates. One of the unique reliability features of the bullion server (among others) is its central memory management.
Memory protection mechanisms guarantee up to 100% memory reliability on bullion. Over and above traditional memory correction mechanisms, such as ECC memory, which maintains a memory system effectively free from single-bit errors, bullion provides much more sophisticated mechanisms such as DDDC (Double Device Data Correction), which corrects dual errors.
The commonly available ‘DIMM sparing’ is now being enhanced to provide rank sparing. With rank sparing of dual rank DIMMs, only 12.5% of the memory capacity is being used to enhance the Quality of Service. If, for example, bullion servers are equipped with 32GB dual rank DIMM memory kits, each kit consisting of two DIMMs with a capacity of 16GB, a 32GB dual rank memory kit thus provides 28GB of useable space, while the rest is being used for fail-over, if the level of ECC errors becomes too high.
Another example of a mechanism to improve memory reliability is ‘MCA recovery’, which ensures that memory errors detected are forwarded to the VMware hypervisor, to make sure that the hypervisor does not use this deficient memory address space any more. These two features limit the impact of memory crashes just to the affected VMs, without having to provide the memory DIMMs needed for memory ‘mirroring’.
Finally, for 100% memory reliability, bullion servers allow for the memory to be configured in ‘mirroring’ mode, with data being written simultaneously in two different memory modules. Thanks to the large number of memory DIMMs available in bullion servers and the falling cost of memory, mirroring is finally becoming more accessible. So bullion can support 2TB of memory for each VMware host, even when memory mirroring is being applied.
Cost benefits that speak for themselves
bullion enables hundreds of business-critical applications to be consolidated on a single server: so energy consumption falls dramatically. Typically a dedicated physical application server consumes about 350 Watts, so an infrastructure consisting of 100 business-critical applications would potentially consume 35KW. Running the same 100 applications, with the same or better performance, a bullion architecture of three bullion servers each with eight processor would consume only 7.2KW: close to a five-fold reduction in energy consumption.
bullion also delivers significant savings on licensing costs, with software costs typically related to the number of licenses, which in turn depends on the number of processors present. bullion delivers 30% more performance per processor than standard small-scale servers, so it requires 30% fewer sockets to sustain the same number of applications and users, with the commensurate savings on the number of licenses needed.
Also, with its high levels of reliability, bullion helps to increase the Quality of Service delivered by customers’ applications. So rather than investing in highly featured software to increase application Quality of Service on smaller servers, many customers prefer to consolidate their virtualization farm onto fewer bullion servers, using a standard software set with its lower cost per license. The combined effect of these different elements can reduce overall software licensing costs by up to 70%.
bullion has been designed to bring simplicity back into the heart of the information system, cut TCO and improve business agility.
- Extreme consolidation to cut costs while optimizing and simplifying IT infrastructures supporting business-critical applications
- Extreme performance to boost applications’ capacity while taking full advantage of exceptional and totally scalable power
- Extreme virtualization to implement business-critical applications faster and in total security and energize the business
- Extreme efficiency to free up space in data centers and cut their carbon footprint.