Technology News you can use...

5 Steps to Successful IT Consolidation

#1 Lay out a change and risk management strategy
#2 Develop a plan for resiliency
#3 Test (and improve) branch office performance
#4 Architect a forward-looking infrastructure
#5 Plan a phased roll-out

For most organizations today, the possibility of consolidating IT infrastructure out of remote offices and into the main data center is an idea that has been on the table to cut costs and boost productivity. After all, isn’t consolidation supposed to produce these benefits with nothing more than a small one-time effort of some time and money?

While consolidation can certainly bring a number of benefits to organizations - and by this point you’ve already created the ROI calculations that show a year or less payback period – it will take more than just a Friday afternoon to ensure that your consolidation project is truly successful. As far too many IT managers will tell you, a poorly-planned consolidation project will have your executives screaming, users threatening mutiny, and IT in the hot seat to quickly undo all the effort that went into the project in the first place.

Like most projects, successful IT consolidation requires time, planning, buy-in, and methodical has its own special set of challenges that relate to the fact that you are fundamentally changing architecture. This has wide-ranging implications that cover everything from budgets and capital considerations over the location of infrastructure and perceived impacts on business units.

Step 1: Lay out a change and risk management strategy

Risk management is part of project management 101, but there are some special considerations in the case of IT consolidation. In any IT consolidation project the benefits are most obvious to the corporate entity (immediate cost savings, less strain on IT) and less obvious to business units or remote offices. After all, these other groups may feel like they are losing “their” servers or local tools and will be penalized with reduced performance for no apparent reason. And, since IT is generally tasked to align their own goals with those of the business, this can create conflict and misunderstanding.

Similarly, IT teams located in remote offices may worry about their own ability to deliver on agreed-upon service levels to business units in a new environment where they may have less control of physical infrastructure to fine-tune.

Step 2: Develop a plan for resiliency

One of the primary downsides of IT consolidation is that your users are more reliant on the WAN and the central data center for data and applications. With this infrastructure shift, IT managers need to carefully consider the resiliency of this architecture.

Systems that were previously distributed provided some level of resiliency; all of an organization’s users wouldn’t be immediately affected by an outage of a system. But, with a centralized system, downtime immediately impacts the whole organization. This problem can, and must be, designed around.

First, consider the redundancy of applications, storage and servers. As application servers are consolidated to the main data center, it becomes easier to provide backup systems without breaking the IT budget. It makes sense to implement backup systems that enable immediate failover for mission critical systems. Putting such systems in place, however, will also require new plans for reliably replicating information from the primary to the secondary server, ensuring consistent failover in the event of an incident. These plans will also require an IT department to consider the location of backup systems: should they be located in the primary data center or in a secondary, off-site facility? Off-site redundant systems are useful for recovery from major disasters, but may introduce additional performance problems discussed in the next section.

Similarly, organizations should consider WAN redundancy. With centralized data and applications, users in branch offices will be more dependent on WAN connectivity, and a WAN outage will have a severe negative impact on user productivity.

When considering WAN redundancy strategy, organizations must make a tradeoff between using two different service providers for last-mile connectivity and using one service provider via different PoPs.

The first option gives an organization greater diversity of transport, and therefore provides greater piece-of-mind that remote offices will always be connected to headquarters.

The second option provides a remote office with a slightly lower level of redundancy; if the Carrier’s WAN goes down the office will be taken off-line. Nonetheless, if just one PoP goes down, the office will maintain connectivity. The benefit of this strategy comes in cost and management. Assuming that your carrier is reliable and has a high uptime to begin with, you are only defending against occasional outages. If a small amount of downtime is acceptable, this approach provides for easier IT management (and one less vendor to deal with), and probably lower costs, since you can buy bandwidth in volume.

Step 3: Test (and improve) branch office performance

There is no doubt that one of the most vocal constituents during this project will be end-users in remote offices. Aside from feeling like they are “losing” their IT infrastructure, they will have serious issues if they cannot access the data and applications they need to do their jobs.

It is important to put this potential problem on the table immediately, and work with end users to test performance. Most users – and many IT staff – will simply assume that this is a bandwidth problem. Given that the consolidation of applications and storage will drive an increased amount of data over the WAN, it is natural for many to consider additional bandwidth to be the de-facto solution to performance across the WAN. But additional bandwidth will often not solve the problem – and in many cases is not necessary. Latency combines with application protocol inefficiencies to create an additional bottleneck to WAN application performance. Many organizations have turned to Wide-area Data Services (WDS) to simultaneously address bandwidth and latency constraints. WDS products combine data reduction, transport protocol optimization, and application protocol optimization to provide WAN application acceleration and bandwidth reduction to all applications that run over TCP.

Often times, organizations just attempt to consolidate infrastructure and plan to deal with performance problems afterwards, hoping that performance will be “good enough” for distributed users. This is a path that will lead to dissatisfaction at best, and mutiny at worst. In fact, many organizations have started a consolidation project only to completely undo it when an executive experiences the pain of working over a congested, latency-impacted WAN connection in a remote office, or when users have just become so frustrated that they demand a reversion to the infrastructure of old.

Why consolidate?

The reasons to consolidate IT infrastructure, including email, applications, and databases as well as tape libraries, are wide-ranging. While most IT managers really focus on just the cost of the physical IT infrastructure, there are far greater benefits that can be derived from IT consolidation:

• Reduced IT management overhead.
• Eased revision control.
• Ease of implementing data security.
• The ability to scale systems more quickly.
• Ease compliance.
• Reduced server and software costs.

Step 4: Architect a forward-looking infrastructure

This is the step where IT departments actually save money. There is much to be saved by moving servers to virtual machines in a centralized data center and consolidating NAS boxes to improve utilization.

The consolidation strategy gives IT the opportunity to scale and secure data more scalably in the future. After all, if the bulk of infrastructure is just located in one place, it is easier to deploy, manage, and test in the case of future modifications.

However the challenge to doing so will be in the technical dependencies of various components in your architecture. How will these changes affect dynamic links to other applications? Does it make sense to move certain groups of applications at the same time? What about backup routes within the network configuration?

This phase of the consolidation process is a good place to get key team members of your application, storage, networking, and support groups (if they are separate groups in your organization) on the same page. Clearly, many of the dependencies discussed in this section cross typical organizational boundaries. As such, it’s important to consider the ramifications of consolidation to your infrastructure and IT processes with all of the team. What applications and technology can be consolidated, and to what degree? Will you need different shifts to handle IT monitoring or user support? How will plans for evaluating, testing, and rolling out new technologies need to be modified?

Additionally, if you decide to consider a WDS appliance for application acceleration across the WAN, you will need to do additional architecting around their implementation. Key to deployment will be exploring various options of where these appliances will live in your network, how they will be sized (various vendors have different approaches), if they will provide some support for disconnected operations in the event of a WAN outage, and if these devices will need redundancy as well.

Finally, loop the earlier WAN redundancy discussion back into your architecture planning. Will you depend on the WDS device for support in the event of a WAN outage, or will you implement some level of WAN redundancy?

Step 5: Plan a phased roll-out

Like step one, this step applies to most complicated technology initiatives within an enterprise. In considering a phased rollout, explore a few different factors:

• How well can you support the offices in the initial phases of the project?
• Which applications make the most sense to consolidate first?
• How complete will the consolidation process be?

The initial phases should focus on offices that can be easily supported. They may be the ones with on-site support staff or relatively close to your IT hubs. In this manner, any difficulties can receive immediate assistance. Communication is easier, too, when these locations are in similar or the same time zones. Another way to slice the phase is by size. Smaller offices may be first on the list, even though they are not the easiest to support, because in the event of difficulties as few people as possible will be affected.

Next, consider the applications you have distributed and which ones it makes sense to consolidate first. Many will go with the “tried and true” consolidation targets like file servers, exchange servers, MS-SQL databases, and web applications. That is a good approach, as it allows organizations to lean on what others have successfully done in the past. Other organizations will often go with less widely-deployed systems, such as document management servers, tape backup systems, time and billing applications, and accounting packages. This is an acceptable approach as well; what is important is that you have thought through the issue with your team and come up with a reasonable plan.

Finally, think through the level of consolidation that will actually occur. While every organization would love to remove everything from a remote office, just leaving a WAN connection, router, and a WDS appliance, that is not always the reality of the situation. For certain offices, full consolidation is fine. But others may find that is necessary to have servers for “scratch” data – that is, temporary files that don’t necessarily need to be accessed by other offices. This could be the case for offices that are quite far from the data center, and shouldn’t have to send all this temporary information over the WAN.

Other offices may be so large that it makes sense to leave some infrastructure there. After all, they may have support staff and such a high utilization of certain applications that they would essentially require the full resources of an application server. In that case, the benefits to 100% centralization do not apply,

Conclusion

There are many benefits associated with IT Consolidation, but it can be a very challenging process. It is key to consider all components of an IT consolidation project before embarking on it. Organizations that spend time and effort communicating the consolidation plan, testing and ensuring application performance, and implementing redundancy, will see tangible results.

If you need assistance or information that will help you make decisions about the future of your business network infrastructure feel free to ask one of our courteous and knowledgeable network experts. Give us a call 1-866-873-3368

Learn about WAN Acceleration: watch Jack's story...