| EDITORIAL |
|

|
 |
Béatrice Piquer-Durand
Marketing Director
Ipanema Technologies |
|
|
Wishing you a great 2010!
|
|
I’m taking advantage of this newsletter to wish you a wonderful year in 2010 on behalf of the entire Ipanema team.
2009 was sometimes good, sometimes grim. May 2010 satisfy our every whim!
2 times the current budget
0 network issues
1 good WAN governance
0 user complaints
Economic recovery or not, each centime spent on the information system must serve to deliver the performance expected by businesses to achieve the company's objectives.
Hervé Hollier, Infrastructure and Operations Director at Bull’s IT Department explains on this page how Ipanema allows him to make major savings while increasing the productivity of his network users.
On page 2 of this newsletter, we will cover a new issue relative to the development of IT architectures and applications "in the clouds": the unified [VPN+Cloud] networking.
Thanks for your trust, and let’s kick off a fantastic 2010 for all of us!
Béatrice
|
|
 |
SUCCESS STORY |

|
 |
Hervé Hollier
Infrastructure and Operations Director
at Bull’s IT Department
|
 |
|
|
Bull reduces loss of bandwidth on its network while cutting telecom costs
|
|
Béatrice Piquer-Durand for Ipanews: Hervé Hollier, could you describe Bull Group in a few words?
Hervé Hollier: Bull’s mission is to help businesses and administrations build, optimize and operate their information systems (IS) while ensuring their cost efficiency, in accordance with stringent security rules. Founded in 1933 and listed on the stock exchange, it generates revenues of more than 1 billion euros. The 8,000 Bull employees work in more than 50 countries worldwide.
BPD: Why did you require an application traffic management solution?
HH: Our rapid development significantly increased the complexity of the network. The number of service suppliers, business applications (IP Telephony, Instant Messaging, etc.) as well as the number of sites grew strongly. Consequently, the increased pressure on resources and the time needed to manage the complexity of the network and traffic were detrimental to the service provided to our production plants. Therefore, we decided to deploy a global and dynamic network management solution.
BPD: What issue(s) did you face?
HH: Users repeatedly complain of various breakdowns and network failures, but we had no way to identify the causes, let alone resolve them. It became apparent that we needed a performance monitoring and optimization tool for our applications if we were to regain the trust and satisfaction of our end-users.
BPD: What criteria led you to choose this tool?
HH: We benchmarked several solutions before choosing Ipanema. We found its rich functionality, multiple dashboards and its capacity to dynamically adapt to changes especially appealing.
BPD: How is the Ipanema solution deployed?
HH: We have currently deployed some 10 physical ip|engines (including one entirely dedicated to measuring Internet traffic). The other sites are monitored by cooperative virtual ip|engines - which allowed us to optimize the solution's acquisition costs!
BPD: What are the benefits?
HH: Since we’ve been using the Ipanema solution, we’ve reduced bandwidth loss on our network. We’ve already recovered more than 30% of the bandwidth without even activating compression and assigned it to the most critical applications. This avoids costly bandwidth upgrades, since the current bandwidth is sufficient for supporting our business activities and ensuring our productivity.
Before, it was always touch and go. Ipanema has allowed us to map our network traffic and gives us insight into the reasons for the failures we experienced. We are now in a position to run our network optimally and to anticipate any future changes. The IT team selected reports that are distributed monthly. The information is available via a Web server shared by the 15 concerned users.
Lastly, the response times of certain business applications have dropped from 30 seconds to one second! As you can see, since we’ve been using Ipanema, productivity has improved significantly, and our users really appreciate it.
|
Download the case study on Bull in PDF
|
|
 |
SOLUTIONS & SERVICES
|
|

|
 |
Thierry Grenot
CTO
Ipanema Technologies |
|
|
[Cloud + VPN]: towards successful cohabitation
|
 |
|
Béatrice for Ipanews: Thierry, is “Cloud” really something new?
Thierry Grenot: It’s not new. Businesses have been using it for a long time: for decades, it's been their VPN that allowed their application servers to communicate with their remote terminals. But when you use applications in “Software as a Service” mode, or when you deploy virtual datacenters to host your business applications, you end up with two clouds: the “plain old” private cloud and the "virtually private" cloud: or just “cloud”. Communications across these clouds are more and more crucial to users who do not care about technical considerations. For them, it is straightforward: there’s one information system and one network.
BPD: What are the consequences of this unified [Cloud + VPN] networking?
TG: The situation must be considered from two main viewpoints:
- The network viewpoint – Access to a branch office can be unique: both for VPN traffic and for the applications hosted in the cloud. Therefore this shared resource must be properly controlled and allocated. The branch can also be multi-connected: the internal and external datacenters can be reached from each access. It’s a very efficient, robust configuration, but it requires control of access sharing as well as selection of the optimum access in order to guarantee the best performance and availabilty for the service, anytime.
- The applications viewpoint – Some applications must absolutely work perfectly well, in any circumstances. This is particularly true for business transactional applications and applications related to customer relationship. The growing cannibalization of the network also needs to be considered: some studies show that half the traffic is recreational, compared to just 15% for business critical applications (OVUM). We believe that this trend will in fact be amplified with the increasing richness of contents and the arrival of the famous Y generation in enterprise.
BPD: How can this traffic be controlled?
TG: While with private networks, traffic can be managed from the datacenters (e.g. with Ipanema’s “virtual ip|engines”), this is by definition impossible with cloud computing, since the goal is precisely to relocate servers to centers that companies no longer control. The problem must therefore be dealt with from the branch office, through:
- Clear understanding of application traffic, including encrypted Web flows used to access applications hosted in SaaS mode;
- Ability to adjust in real-time resource allocation to the instantaneous demand according to the business criticality and the technical requirements of each application flow;
- Dynamic switching of flows, enabling maximized performance and minimized costs through optimum use of available network resources.
All this must be performed automatically, to relieve the technical staff. Finally, the total cost of ownership (TCO) of such a solution must remain marginal compared to the cost of the application services whose performance must be guaranteed.
BPD: What is your preferred approach and the resulting benefits ?
TG: WAN governance, which consists in managing the global traffic from a small number of application SLAs, i.e. performance objectives described in natural language. These objectives are used by an Autonomic Networking system that can adapt to each circumstance, the way a driving aid system adapts to road conditions and the driver's driving style. The main benefits for the company are:
- Optimum performance for all applications, whether they are hosted in its datacenters or in the cloud;
- Improved application availability, here again regardless of where they are hosted;
- Streamlining of network operations;
- Intuitive management, using KPIs that are easy to understand and that can be shared with management;
- Lastly, significant savings (network, operations, servers and user productivity).
|
|
|
 |
|
 |
 |