Service organizations that develop a robust portfolio of non-project services also benefit from creating customers-for-life, which enables not only involvement in a larger percentage of client IT portfolios on average, but also ensuring that the VAR’s business is well-positioned for follow-on hardware and software sales due to client intimacy.

There are two primary objectives that need to be achieved to growing services with enterprise customers:

• Increase the number of your customers who have a service relationship with you.
• Increase the total amount your customers spend annually on services.

In support of these primary objectives, an organization can achieve significant progress by focusing on three characteristics of expert service delivery:

• Sales excellence.  Services sales are different from product sales—they’re longer, more strategic, and if done successfully contribute to long-term, durable relationships.  It is services sales that allow enterprises to purchase a total solution.
• Customer relationship excellence.  Customers can be retained for longer and spend more when relationships are invested in.  Durable customer relationships also provide strategic feedback that enables a business to refine its “solutions” to the specific needs of customers, ultimately shortening sales cycles and improving profit margins.
• Delivery excellence.  Substantive expertise in architecting, implementing and monitoring critical enterprise systems and infrastructure immediately establishes trust and credibility.  Delivery excellence also produces real results for clients in the form of tangible business value.

A significant number of VARs are attempting “managed services,” the latest buzz term, which are intended to provide a high value service for client companies where IT operations aren’t the core of their business.  Additionally, managed services provide a constant connection to each client, which allows a higher degree of engagement in solving IT problems for each client enterprise.  “One big area of investment in 2010 will be managed services, where VARs plan to increase their already growing business. In 2009 (sic), managed services accounted for an average of 9.5 percent of revenue, up from 8.6 percent in 2008. That percentage is expected to jump to 11.4 percent next year.” (CRN Study:  By Scott Campbell, ChannelWeb, Dec. 14, 2009)

While it can be fairly straight-forward to package and market services, the sales process is generally longer.  Additionally, service packages that don’t have significant substance behind them are less likely to create significant revenue.  The sophistication of processes, controls, reporting and ongoing improvement all of these elements, aligns with the level of strategic fit and budget allocation at each client. 

Strategic importance.  More budget.  This is the substance and reward of being a true value added reseller.  

-Chris Westfall

• Consider RTO and RPO:  Selection of a data replication method should start with a business impact analysis to determine required recovery time objectives (RTO) and recovery point objectives (RPO). These will help define the data constraints.

• Choose Synchronous/Asynchronous:  For applications that can’t accept data loss (RTO equals zero), synchronous replication is required. Latency in synchronous replication is an important consideration, because it will drag down application I/O performance. If there is any risk of latency or unreliable bandwidth, consider asynchronous replication. Similarly, for replication across greater distances (50 km to 300 km for example), go with asynchronous replication.

• Enable Application Performance:  Application performance can be impacted by the replication platform. Take into account that host-based replication competes with applications for valuable processor, memory and I/O resources.

• Manage Bandwidth:  Have a clear understanding of the bandwidth requirements, impact on bandwidth cost, and how data replication will impact other applications and users. Clearly understand and take advantage of replication features related to bandwidth, such as compression, bandwidth throttling and configurable bandwidth usage depending on the time of day. Wide-area network optimization devices will help to preserve bandwidth.

• Assume Heterogeneity:  Replication products that support heterogeneous environments can substantially reduce cost by supporting less-expensive or legacy arrays. They also limit vendor lock-in.  Additionally, any disadvantage of array-based replication is offset by the advantage of close integration between replication and the storage platform (and easier support), eliminating the risk of finger-pointing in multivendor configurations.

 - Pete Smith, Consulting Principal

Fundamentally the agile method focuses on teams working in a highly collaborative fashion, communicating frequently, breaking work into bite sized tasks, accommodating change, and continually producing high value results (capabilities, functionality, etc.). 

In the context of IT project management for an enterprise customer I’ve found that there are a core set of agile practices that work well.  These include:

• Daily stand-ups–first thing every day–getting the cross functional team in a room for 15 minutes to quickly review who’s doing what, what obstacles have come up, how to get them out of the way, and coordinate hand-offs (implementation to testing for example).
• Dissect the ‘end product’ or final objective into bite-sized pieces.  In agile lingo, “stories” should be created for all aspects of a project–every element of design, development, implementation, testing, and acceptance.  Basically, break the project down into pieces that your parents can understand.
• Tackle issues continuously.  Change is a constant, not an exception, so focus on it and treat change as standard process, so that reviewing and adjusting are managed proactively.  I’ve found that because “stories” are easily managed units, the ripple effect of change is easy to handle.
• Thorough documentation.  I don’t mean binder upon binder of verbose prose.  I mean tracking every piece of work and documenting it in ways that matter to each party involved in the process–from producers and implementers to the customers.  Transparency is the key.
• Fanatic communication. Agile is all about collaboration to create efficient forward progress. Communication is essential–not only in daily stand-up meetings, but also between team members, and across ‘functional’ groups, and certainly between vendors and customers.

The results from this approach can be really great. 

• Customers know what’s happening and business alignment is more certain.  Keep in mind that you have to get your customer on-board with the agile method to begin with.  It is probably different than what they’re used to.  It’s certainly not just setting a massive deadline and scoping to the finish-line.
• From a project management point of view, another great outcome of an agile approach is that vendors and other groups within IT are tightly coordinated, and usually highly motivated because this highly collaborative approach gives everyone a stake in each successful milestone. 
• Finally, incremental progress is delivered continuously.  Put another way, it’s possible to eliminate risk by not focusing on one massive deadline, but rather lots of small deadlines.

There’s certainly a lot of moving parts to any large-scale IT project, and an agile approach to managing the project won’t change that. But adopting an agile method can create some great results.  A strongly committed and collaborative team certainly increases efficiency and contributes significantly to producing high quality outcomes.

 - Chris Westfall

NAS clustering is defined by using distributed file systems running concurrently on multiple NAS nodes. The clustering provides access to all files regardless of node location.

NAS sprawl becomes a problem as file creation increases and file size grows. To tame sprawl, virtualization and data protection need to be implemented. This is where NAS clustering comes in.  NAS clustering pools physical storage from multiple NAS to act as a single storage device that is managed from a central console. Better data protection can be achieved through this virtualization because it enables a singular point of control of all the data.

It’s important to consider how high availability resolves issues with data growth and management. Here’s how:

• NAS storage is easy to grow and there are no actions to be taken on client systems.
• NAS storage allows IT to linearly scale growth to many nodes.
• Aggregation of throughput and IOPS are node independent.
• Management becomes centralized due to the reduction of separate NAS servers.
• User and application access is simplified with load balancing built in, and data protection and replication are also part of the architecture.

Finally, it is important to anticipate and plan for potential issues that arise when implementing clustered NAS.  For one, clustered NAS cannot be implemented with legacy NAS servers; all data would need to be moved to the clustered environment.  With this in mind, it’s best to implement clustered NAS when implementing a new NAS solution. Additionally, as always, it’s important to plan for the right hardware and licensing so that costs are anticipated and managed.

-Pete Smith, Principal Consultant, Signal Creek Technology

Signal Creek has helped many organizations plan and implement DR, and our experiences suggest six key steps to align, implement and test for Disaster Recovery and Business Continuity:

Step 1:  Business Alignment

The first step of any good disaster plan is extensive planning.  Many times this planning originates at the IT level, which can be misguided.   Planning should begin with the business owners.  Only the business owners in an organization can determine which information is critical and needs protection, and the value that this information has for the organization.  Based upon meetings and consultations the following, at a minimum, should be defined.

• What type of information is vital to the continued success of the organization and therefore must be protected at the Disaster Recovery site?   A formal classification process of all information by importance is always encouraged during this process.
• What is the Recovery Point Objective for this information?
• What is the Recovery Time Objective for this information?

Step 2: Inventory Assessment

The next phase for Disaster Recovery is to match the requirements defined from step 1 to existing IT resources and technologies to see what can be accomplished with current systems and personnel.  For large organizations with a multi-disciplined IT staff this process may be handled internally.  For smaller organizations it may be best to consult with external resources so that the best technologies and solutions can be matched to the business requirements.   Once this process is complete a gap analysis should be created to identify where new products or solutions need to be procured to meet the defined business recovery requirements.

Step 3: Solution Procurement

Often the gap analysis established in step 2 sets the foundation for a formal request for proposal process.  RFPs can vary significantly in size and complexity depending upon the significance of the application to be protected and its underlying cost and complexity.  Don’t rush this process.  You want to make sure that you find the right solution at the right price.  There are many solutions out there and not all of them do what they say they do.

Step 4: Implementation

You are finally ready to implement the solutions you have decided upon.  Project management is highly recommended for this step of the project.  Many projects get stuck in a continuous implementation cycle when hard deadlines, proper resource allocations and formal scope change processes are not in place.  The project manager should always be involved as early as possible in the Disaster Recovery planning process.

Step 5: Documentation

A Disaster Recovery plan cannot be successful without complete and accurate documentation.  If external vendors are involved in the process make sure they deliver adequate documentation prior to completing their phase of the project.  The documentation should not just be technical in nature.  It has to be designed to support recovery of the business by personnel who were not involved in the original Disaster Recovery project implementation and design. It’s critical to build clear documentation and then evangelize it with all key employees.

Step 6: Test

Assume that you will not be successful in recovering from a disaster if the solution has never been tested.  Ask anyone who actively tests their Disaster Recovery plan and they will tell you that although testing is painful many unforeseen and unaccounted for situations arise that would have resulted in a DR failure if an actual disaster had occurred.   Performing Disaster Recovery testing can be an extremely complicated and technically challenging process especially when it is only considered after a Disaster Recovery solution has been implemented.  Testing requirements should be considered during all aspects of DR planning and implementation.  And beyond implementation, Disaster Recovery testing should occur at least once a year. Often different components are tested at different times during the year so that normal business operations are not too adversely affected by the testing process.

As a whole, the above six steps are meant to provide some high-level milestones necessary to achieve successful Disaster Recovery.  Let us know if you have questions, comments or need some guidance.

-Andrew Miller, Principal, Consultant

Storage can play a major role in creating a Green IT infrastructure, but it must be carefully thought out and planned. For example, modern disk drives are thin and high-density. The drives are smart. They are designed to deliver “just in time” performance, by spinning up to the right location at the right time. Improved capacity and performance means that we can pack hundreds of terabytes of data into a rack.

This dense rack offers unique challenges for cooling and power management. In the last year I have seen an array of disks loaded into a rack use more than 30KW on multiple occasions. A five year old data center was most likely built out to efficiently handle 10KW to a rack. Green Storage is great, but it must be monitored and managed. My short list of items to measure would include:

• Air pressure
• Air particulates
• Leaks
• Electricity levels
• Temperatures

Monitoring these levels will help you reduce your exposure to power and HVAC outages and help you understand the energy impact of your Green Storage. If you want to get started, try the free tools offered by The Green Grid (http://www.thegreengrid.com).

The Green Grid is a global consortium of IT companies and professionals seeking to improve energy efficiency in data centers around the globe. Here are a couple of tools you can soon download from their site:

1.    Power Efficiency Estimator

The new Power Efficiency Estimator is designed to help data center operators compare different scenarios of power topologies and technologies inside of their facility. It takes factors such as workload, availability, and space constraints into consideration and generates a report that is designed to improve decision-making in a facility.

2.    PUE Estimator

The new PUE Estimator allows data center operators to easily input their facility’s specific data at regular intervals to determine their PUE. Power Usage Effectiveness (PUE), a metric created by The Green Grid, determines the amount of energy used by the facility and the IT gear inside of it.

Armed with these tools and basic wireless monitoring devices strategically placed in your data center, you can continue to effectively reduce the amount of physical storage required, cut down on the carbon footprint and lower costs.

This type of offsite storage and computing has been around for awhile, but has only been available to companies that could afford high speed networks, which make accessing remote data practical. With high speed networks becoming ubiquitous and affordable, a wide variety of businesses can and should consider Cloud Storage as an alternative to in-house storage systems. 

Let’s look at the advantages and disadvantages to Cloud Storage.

Advantages:

• Data access is possible from any location that has a connection to the internet.
• Implementation can be simple.
• Expansion is easy and efficient.
• Many parties can be granted access to your data.
• There’s no need to physically manage storage (fix bad disks, backup data) if the storage is bundled as an outsourced managed service.

Disadvantages: 

• Performance can suffer with internet traffic. Remote storage will never be as fast as a local, SAN-attached storage. Also, any shared resources may be competed for by other customers.
• Security is a concern. When a company’s sensitive data is transported over an unsecure network such as the internet, there’s a potential risk of exposure.  But, even a corporate network has its risks.  Exacting security practices can alleviate this concern.
• Cost can rise dramatically if usage is not monitored closely. Out of sight should not be out of mind; Cloud Storage should be proactively monitored.

Based on our experience at Signal Creek with a range of our partners, the state of Cloud Storage today seems most advantageous to smaller businesses, as opposed to the larger organizations we typically work with.  With the heavy demands of large enterprise storage and archiving, Cloud Storage isn’t a clear choice for large organizations with enormous amounts of data, or heavy data security and compliance requirements.

- Pete Smith, Principal, Consultant

For all of us in the computing industry, it’s a throughput problem. All of the cars will fit on the highway (even the Hummers), but everyone will crawl along at 20 miles per hour. I have seen this same throughput problem in the Data Centers of customers who have virtualized their environment, without proper planning.

To be sure we share a common definition:  A virtual server is a piece of software that sits on a physical server and can run different operating systems and applications as if it were a physical computer; it even has virtual CPU’s, RAM hard disk and NIC cards. The operating system doesn’t know it is running on a virtual server; like HAL in 2001 a Space Odyssey, it thinks it is “real”.

Technologists have been quick to grasp the inherent advantages of a virtual machine.

From a storage perspective virtualization allows you to pool resources, increase utilization and simplify maintenance and management. The problem can be traffic – or more specifically I/O. Heavy workloads will have an impact on performance. When the engineers at Signal Creek Technology design an enterprise Data Center we follow best practices to address this capability issue as well as capacity.  A proper capability analysis will allow the Storage Administrator to move less critical data to less expensive drives such as SATA. The costly high-fidelity drives (such as Fibre Channel) can be reserved for the most important applications and databases.

A good design would include a review of the physical storage environment before virtualization to assure that multiple applications running on multiple operating systems don’t overwhelm the throughput capacity of the available storage. We look at the aggregated workload and peak demand to properly measure application I/O. Only then can you determine how much storage you need to handle your most critical applications.

In the Data Center there is a constant struggle with under and over utilization scenarios.  Large online retailers such as Amazon and eBay have been able to use virtualization technologies to create marketable excess capacity in their storage environments. They sell this capacity to other companies who need more storage. But for an enterprise that doesn’t leverage capacity as part of their business: Under-utilize a storage resource and you waste corporate resources. Over-utilize and you risk load balancing and application errors. Virtualization allows the Storage Administrator to manage these scenarios in an efficient manner.

Understanding your Data Centers throughput and capacity will help you avoid the biggest traffic problems and assure the success of your virtualized environment. If you’d like help designing a virtual Data Center, or have questions or input for us, email expertservices@signal-creek.com.

 -Matt Hamilton, Principal, Consultant

Typically the only option available to an end-user is to export or archive their Outlook email to a PST file.  Before you know it you have PSTs on network file shares, PSTs on local hard drives, PSTs everywhere!    Good luck backing these files up, let alone trying to satisfy any legal or regulatory requirement to search email, including PSTs.

This is where Enterprise Vault comes in.  Enterprise Vault provides a variety of mechanisms to find, import and remove existing PSTs, whether they reside on network shares or client computers.  Enterprise Vault can also disable Outlook Auto-archiving so that new PSTs don’t get created.  You can also use Group Policy settings to prevent new PSTs from being saved.  Enterprise Vault offers a few different tools to find and import PSTs.  The two that I am going to discuss are the Client-Push method and the Server-Pull method.

Once Enterprise Vault is up and running and you are actively archiving users’ email, you are now ready to tackle eliminating PSTs.  Typically the place to start is with the Client-Push method.  The Client-Push method utilizes the Enterprise Vault Outlook plug-in to “push” email data to the Enterprise Vault server.   This method has the advantage of being able to archive PST’s while they are open and will prompt the user for their password if the PST is password protected.

Once you have archived PSTs with the Client-push method, there will undoubtedly be others that the pull method can address.  The Server-Pull method pulls PST data from client computers or servers.   This method is useful when you have a server or network share that contains lots of PST’s that are rarely accessed.  The Server-Pull method finds computers on your network and then searches those computers for PST files.  Once the find process is complete you can begin the process of archiving the contents of these PSTs to existing Enterprise Vault user archives.

Both of these methods allow you to create shortcut stubs to the archived item in the original PST file or back into the users Outlook mailbox using a folder structure that you create.  You can also choose to just archive the items and remove the PSTs from their original location.  The users would then use one of the many Enterprise Vault search tools to search for archived PST email.

If you would like to learn more about Enterprise Vault and PST archiving, Symantec has some great literature on their web site to help you get more familiar with the product.  And of course, you can always get hold of us at expertservices@signal-creek.com with any questions or for some help.

- Andrew Miller, Principal, Consultant

Historically DD was a feature of appliances such as VTLs and WAN optimizers. Fortunately for many companies that cannot or do not want to afford the costs of these appliances, DD is now being incorporated into backups applications.

Some key, specific benefits of DD include:

• Companies can see benefits in lower storage space requirements that will save money on storage budgets.
• In terms of storage space alone, we’ve seen up to 90% disk space savings. 
• The more efficient use of disk space also allows for longer disk retention periods, which provides better recovery time objectives (RTO) for a longer time and reduces the need for tape backups.
• Data deduplication also reduces the data that must be sent across a WAN for remote backups, replication, and disaster recovery.
• Reductions in power, space and cooling requirements are also a benefit of DD.
• Within the virtual environment the ability to deduplicate VMDK files is a big benefit.

The major vendors have DD capabilities, including EMC, Symantec, IBM, CommVault, Quantum and the like.  Check it out.