Blog Host: Brett Cooper READ ENTIRE BIO

The future is based on evolution
Many organizations look at their current environments and realize that there are challenges to overcome; from improving their data protection capabilities to making their infrastructure more flexible to deliver on their promise to provide solutions to their end customers' needs. In each of these discussions, the idea is that the vendors and the partners in technology will continue to improve on their products. The reality is that all of these improvements come as a result of a symbiotic relationship between customers and vendors. Vendors deliver solutions based on the customer requirements.

At one point in the technology space, there was a true research and development (R&D) department within most vendor organizations, but today the research piece has all but disappeared and the development arm has taken over; most organizations cannot afford to research solutions that may or may not have any revenue implications. The development arm of the organization -- or as many call it, "engineering" -- is tasked with delivering solutions based on real-world customer requirements that can deliver revenue to a vendor in a specific period of time or event horizon. In other words, most vendors don't invent technologies that they think are cool, rather they invent technologies based on a customer saying, "We have this challenge (problem). If you can solve it, we will buy it." This evolution of technology has driven the industry for sometime and will continue to help customers and vendors keep fueling the future of the solutions. Look to this process to be one that is based on baby steps, not gigantic leaps into the future. Unifying the storage and storage management space with solutions and integrating these solutions into the current infrastructure of the organization takes time. No one really wants to do a forklift upgrade as there are larger implications to the organization that need to be considered.

Tighter integration of applications and storage
I believe that storage does not exist without an application. No one goes out and buys a hard drive or storage array without a purpose in mind for using the space to hold some kind of data. It is only logical then that if an application were tighter integrated into the storage environment, then an organization could get more value out of that storage environment, whether it comes from the reduction in management costs or the ability to increase availability.

The next frontier for storage is closer then you think
You may be reading this article from the comfort of your home computer and wondering, "How can all of this next generation technology help me?" I believe that many vendors are looking to deliver storage solutions directly to the home that take the features from their larger platforms and offer them to smaller organizations that do not have a IT staff available.

I have long wanted to have a storage appliance in my home so I can store all of the digital media that I have accumulated over time and share the collection with my family. Also, my children have a great number of educational games that they play on a regular basis and they are not as careful with CDs as they should be, so I want to copy them to a shared storage appliance and run a CD emulator package (licensing is still OK, as I am only using one copy), so they don't destroy the CDs and can still play their games. I could go out and build a host that has dedicated storage behind it along with data protection software, but this could take me many weeks to do. This is what many smaller organizations in the market are doing. I could also upload all of my digital assets to a broadband storage provider and then view and download only what I need, relying on the provider to deliver the data protection and other features that I require, paying a monthly or yearly fee for the service; however, like many others, I am a bit of a control freak and want the storage locally in my home.

The storage would need to provide a file protocol support, low power and heat requirements, as well as to deliver data protection and high availability (RAID) capabilities, all in a low-cost package that is backed by great service and support. I am thinking the $1,000 range with multiple hard drives along with software that supports my environment for management. I have seen many solutions at my local electronics stores, but none of them have been able to deliver what I need for my environment. I believe many of the larger enterprise storage vendors could not support this package since their service and support models are very costly. Rather, I think, they would need to spin out or OEM the solution to another vendor. Think of it as the Cisco and Linksys solution to the storage equation. Cisco provides the enterprise networking gear, while Linksys is squarely positioned to provide the Small Office Home Office (SOHO -- another great acronym) networking solutions. This could be a space for an EMC or Network Appliance to deliver a compelling solution to the market.

What do you think?
Posted by Brett Cooper Storage protocols: Who cares about the protocol?
30 JUN 2005 23:48 EDT (03:48, GMT)
For years the debate has raged over the right protocol to use: block (Fibre Channel or iSCSI) or file (NAS protocols such as CIFS or NFS). Having watched this debate while working for a variety of vendors I have always found it interesting to hear some staunch supporters of each ring in their support by disparaging one benefit of their protocol of choice only to have a supporter of the other side of the debate fire back with what they believe is a better feature of their side of the protocol debate. To borrow saying from Susan Power, the blonde, short-haired woman in the exercise and health videos, "Stop the insanity!" The protocol of choice is usually dictated by a couple of things:

1. The application being used. Many applications specify which protocols they support based on testing. For Microsoft Server-based products, such as Microsoft Exchange and Microsoft SQL Server, the only supported protocols are blocks-based, unless the protocol is being run on a Microsoft Windows Storage server platform (OEM'd by many of the leading vendors in the market today). Then there is support for the CIFS protocol, up to a certain performance level. With this said, it is important to check with the application vendor, OS vendor and your storage provider to insure that there is support across the full solution for the protocol that you have selected to run your application on. In the end, it is the application that is critical to support with the specific protocol that will be in use.

2. The expectations of the organization. Some would say that history usually tells what the future will bring. I believe that this is true a certain percentage of the time, so look to the standards list within the organization for guidance based on what has been deployed in the past. Most data centers in the world are running blocks protocols, mainly Fibre Channel, for their most important applications, while some are running NAS protocols. The reason one was selected over the other has to do with perceived ease of management, performance and maturity. Look to support the full mix of the solution within the organization and spec out the various parts of the solution to integrate with what is already on the floor.

3. Requirements for data protection for the organization. This part of the discussion is one that I find happens usually after a solution is installed within an organization, but should be discussed beforehand. File protocols have the ability to snapshot the data on the storage system, usually without integration into the higher level application or OS platform, as they control the file system and the cache on the storage system itself. This provides the ability to coordinate snapshots and other data protection capabilities within the storage system, while blocks protocols usually require integration into the application and host OS platform to get consistency of the file system. In either case, most storage vendors provide solutions that integrate into the host OS and application to deliver the same capabilities to a block solution that are available in a file solution. Look to understand what the data protection requirements are of the organization when looking to a protocol-based solution and include any requirements for mirroring replication within the discussion.

4. And some would say price. I always have taken issue with the pricing discussion as I believe, along with many analysts (not that I am an analyst, but am using their wisdom to support my position), that the initial purchase price of any solution is just that: the amount of capital cost that needs to be depreciated over a given amount of time based on the organization's accounting rules. While this is an important budgetary number, the ongoing price of managing the environment is also important to understand, as this budgetary amount is one that will continue into the future on an ongoing basis and may be some 85% of the total cost of the solution based on the life of the solution within the organization. (The 85% number came from an analyst report I read several years ago.) So, if this analyst was even 50% correct on his or her numbers, the total price of the solution is in favor of understanding what the management costs are in addition to the purchase price to understand the true total cost of ownership (TCO).

I have had the benefit of working with some very intelligent individuals who were responsible for coining acronyms that have taken the industry by storm. These individuals had the forethought to integrate the real business value of their offerings with a sensible name. I have also seen some acronyms that have been less than well thought out. I won't point any specific ones out, but I am sure that they will be the fodder for some B-school lecture on the mistakes in marketing at some point in the future.

I personally think that all of the focus on naming is a bit overdone, since the real value is putting the name into real customer English (or what ever language the product is being marketed in). This is where the real value of the product is and how customers want to be marketed to. If it takes 30 minutes to read a data sheet on what a product does, then what is the purpose of naming things to begin with?

I guess I believe in what I call the "Internet way of doing business." Before I click on something, I better know where it is going and why I am going there. If there is a name I am unfamiliar with, I usually don't select it. This relates well to all of the confusing names in our industry as well as the acronyms. My wife once told me that when I am on a conference call with my colleagues it sounds like I am talking another language full of acronyms that only I would know. I have come to find that this is also true in almost every industry; each industry has its own lexicon or jargon that is known within the industry, but not to outsiders, kind of a secret handshake of sorts.

Anyway, if you were wondering what the three acronyms that I started this entry with are, here are the answers and a definition from the Storage Networking Industry Association (SNIA) Web site, in human readable English.

Information Lifecycle Management (ILM):
The policies, processes, practices, services and tools used to align the business value of information with the most appropriate and cost-effective infrastructure from the time information is created through its final disposition. Information is aligned with business requirements through management policies and service levels associated with applications, metadata and data.

Continuous Data Protection (CDP):
CONTEXT [Data Recovery]
A data protection service that captures changes to data to a separate storage location. There are multiple methods for capturing the continuous changes involving different technologies that serve different needs. CDP-based solutions can provide fine granularities of restorable objects ranging from crash-consistent images to logical objects such as files, mail boxes, messages, etc.

Storage Management Initiative Specification (SMI-S):
A specification that documents a secure and reliable interface that allows storage management systems to identify, classify, monitor and control physical and logical resources in a Storage Area Network.

While working on iSCSI technologies at Network Appliance, I had the opportunity to work with the iSCSI standard; before that I worked with Fibre Channel Technologies and their associated standards. All of the protocols support standards that dictate the way an initiator, switch (in the case of Fibre Channel) and target handle the various stages of establishing an active session and transferring data back and forth (reads and writes) and that manage the overall environment (including security). This may seem like an easy task, but for those who have read the various standards that have been ratified by the IEEE and SNIA, as an example, you will come to appreciate the efforts that the various technologists have invested in creating the standards that are embedded in the products we use every day.

One of the latest standards that is having an impact on the products we use is SMI-S. It is being supported by the efforts of the members of the Storage Networking Industry Association (SNIA). SMI-S defines a provider and consumer of information for management in a commonly available and formatted information string (in this case, XML). Think of this as SNMP on steroids and specifically focused on storage and storage management; it includes host-based storage, the interconnects (FC switches) and the storage devices themselves.

To make my point on the importance of support and adoption, I thought I would include a list of SMI-S vendor-supported products as provided by the team at SNIA:

From the SNIA SMI-S Web site:

This portion of the program report lists vendors that have announced products and programs that are in support of SMI-S. The vendor products that have passed conformance tests for SMI-S 1.0 through SNIA-CTP are marked with an asterisk (*). To learn more about what vendor's products that have passed SNIA-CTP please go to www.snia.org/ctp/. If you have a product that supports SMI-S and would like it listed below, please send the information to hope@snia.org.

* HP StorageWorks Enterprise Virtual Array (EVA)	* HP StorageWorks Modular Disk Array (MSA)
* HP StorageWorks Disk Array (XP)	* HP StorageWorks Core Switch 2/64
* HP StorageWorks Core Switch 2/64 PowerPak	* HP StorageWorks SAN Switch 2/16
* HP StorageWorks SAN Switch 2/16 EL	* HP StorageWorks SAN Switch 2/16 PowerPak
* HP Sturestore FC 1GB/2GB Switch 16B	* HP Brocade Silkworm 2800 FC Switch
* Computer Associates BrightStor SAN Manager	* Veritas CommandCentral Storage
Veritas Volume Manager	* AppIQ StorageAuthority Suite
* Engenio Information Technologies E-Series	* EMC Symmetrix Storage Arrays
* EMC CLARiiON Storage Arrays	EMC Celerra File Servers
* EMC ControlCenter Storage Management Software	EMC VisualSAN Storage Management Software
Hitachi Data Systems HiCommand Device Manager	Hitachi Data Systems HiCommand Storage Manager
* Hitachi Data Systems Lightning 9970V	* Hitachi Data Systems Lightning 9980V
* Hitachi Data Systems Lightning 9960	* Hitachi Data Systems Lightning 9910
* Hitachi Data Systems Thunder 9580V	* Hitachi Data Systems Thunder 9570V
* Hitachi Data Systems Thunder 9533V	* Hitachi Data Systems Thunder 9532V
* Hitachi Data Systems Thunder 9531V	* Hitachi Data Systems Thunder 9530V
* Hitachi Data Systems Thunder 9200	* Hitachi Data Systems Thunder 9585V
* Hitachi Data Systems TagmaStore USP	* Hitachi Limited SANRISE USP
* Hitachi Limited SANRISE 9970V	* Hitachi Limited SANRISE 9980V
* Hitachi Limited SANRISE 2200	* Hitachi Limited SANRISE 2800
* Hitachi Limited SANRISE 9585V	* Hitachi Limited SANRISE 9580V
* Hitachi Limited SANRISE 9570V	* Hitachi Limited SANRISE 9534V
* Hitachi Limited SANRISE 9533V	* Hitachi Limited SANRISE 9532V
* Hitachi Limited SANRISE 9531V	* Hitachi Limited SANRISE 9530V
* Hitachi Limited SANRISE 1200	* Hitachi Limited SANRISE 1100
* Sun Storedge 3000	* Sun Storedge 6000
* Sun Storedge 6130	* Sun Storedge 6920
Sun StorEdge Enterprise Storage Manager	CreekPath Suite 3.1
IBM SAN Volume Controller	IBM SAN Integration Server
* IBM TotalStorage Productivity Center	* IBM Enterprise Storage Server
* McDATA Intrepid 6140	* McDATA Intrepid 6064
* McDATA ED-5000	* McDATA Sphereon 4500
* McDATA Sphereon 4300	* McDATA ES-3032, ES 3016
* McDATA ES-3232, ES 3216	CreekPath Suite 3.2
* Dell | EMC CLARiiON Storage Arrays	* CNT FC/9000 Fibre Channel Directors
* Cisco MDS 9509 Multilayer Director Switch	* Cisco MDS 9506 Multilayer Director Switch
* Cisco MDS 9216 Multilayer Fabric Switch	* Cisco MDS 9140 Multilayer Fabric Switch
* Cisco MDS 9120 Multilayer Fabric Switch	* Brocade Silkworm 2010
* Brocade Silkworm 2040	* Brocade Silkworm 2050
* Brocade Silkworm 2210	* Brocade Silkworm 2240
* Brocade Silkworm 2250	* Brocade Silkworm 2400
* Brocade Silkworm 2800	* Brocade Silkworm 3200
* Brocade Silkworm 3250	* Brocade Silkworm 3800
* Brocade Silkworm 3850	* Brocade Silkworm 3900
* Brocade Silkworm 12000	* Brocade Silkworm 24000
* Network Appliance Enterprise Storage System	* SGI Infinite Storage
SGI InfiniteStorage Resource Manager	SGI InfiniteStorage Volume Manager XVM
SGI InfiniteStorage Shared Filesystem CXFS	* StorageTek B-Series
* StorageTek D-Series	HP OpenView Storage Area Manager
* IBM Total Storage FAStT Storage Server	3PARdata InServ Storage Server S200
3PARdata InServ Storage Server S400	3PARdata InServ Storage Server S800
* QLogic SANbox 5200	* Emulex LightPulse FC HBA Family
* HP VA7100	* HP VA7110
* HP VA7400	* HP VA7410
CreekPath Suite	CA BrightStor Storage Resource Manager
* Crosswalk Storage Manager 2.0	* HDS HiCommand Storage Services Manager v4.0
* HP Storage Essentials v4.0	* Sun Enterprise Storage Manager Advanced Application v4.0

I am seeing the adoption of SMI-S outpace all of the managements standards of the past, in terms of adoption and real-world implementations.

Let me know if you have any questions or thoughts on standards and SMI-S.
Posted by Brett Cooper Maslow's Hierarchy of Needs for storage management, part 2
27 JUN 2005 17:06 EDT (21:06, GMT)
On Friday I talked about the levels of Maslow's Hierarchy of Needs as they relate to building a storage management solution, and today I will continue and conclude my thoughts on this topic. You can look back at Friday's entry to understand the levels of the management offering and how they are interrelated, building on one other, so each level advances the capabilities of the storage management solution to deliver a solution to the needs of the organization. Defining the layers:

1. 1st layer -- physical infrastructure monitoring capabilities
2. 2nd layer -- virtual or logical view of the environment
3. 3rd layer -- common services
4. Final layer -- advanced functionality

One of the best practices that I have seen in successful implementations is the creation of reasonable and objective goals that are measurable. I would suggest measuring the goals at certain intervals after implementation: 30, 60, 90, 180 and 365 days. While measuring the results, also look to subjective feedback from your team and the customers that you have, both internal and external to the organization and tune the implementation along the way. One thing that I always keep in mind is that technology solutions are iterative processes, and there is no 100% perfect solution; each has its pluses, minuses and associated risks. The key is to create platforms that work together and can modularly fit into the overall infrastructure and grow as the organization grows, leveraging the Three Ps to deliver the best possible solution based on the goals of the organization.

Let me know what you think of the Three Ps and using Maslow's Hierarchy of Needs to explain storage management solutions.

Tomorrow, I am thinking about discussing standards in storage and storage management. Do you have any thoughts on standards (e.g., SMI-S)? Let me know. Thanks.
Posted by Brett Cooper Maslow's Hierarchy of Needs for storage management
24 JUN 2005 22:48 EDT (02:48, GMT)
Back for the next entry in my blog, and today we are going to look at building a storage management solution using Maslow's Hierarchy of Needs as a blueprint. In my opinion, the goal of a storage management solution is simple: to reduce the complexity of the common tasks associated with managing a storage environment, thereby reducing the costs and the risks of the environment, which, in turn, enables the business to do what it does best.

To give some background and a real-world example of this, consider a company that makes toothpaste and has a good-sized IT department and multiple PBs of storage under management. The company is looking for ways to improve their efficiency in their IT department and increase availability and performance of their storage environment. The CIO of the company realizes that if he can reduce the costs of the IT infrastructure then he can increase the profits by enabling the company to better understand their end customers using the information available to him in his applications and storage environment. He has worked on an equation that allows him to reduce his costs on a per unit shipped number. In this way he has equated his return to the business in a manner that everyone within the business can understand. He simply took the total unit volume that the company planned to ship this year and divided his projected savings and increase of information available to come up with his numbers. This simple equation provided him with the necessary backup to get the green light from the management team to go ahead with the consolidation of his environment and the movement to a new storage management solution. He had help from a bunch of key players throughout the organization, but in the end he made the call to move ahead. As with any larger decision, there were heroes involved on an individual contributor level and several vendors that helped move the decision making process forward.

Now that we have a goal set, we need to leverage our products to create solutions to the challenges of the organization. The top level of Maslow's Hierarchy of Needs talks to "Transcendence," which in our discussion means a higher level of management that was before not possible within the organization. This may be because one of the three Ps was out of alignment or simply because the latest products had not been leveraged to deliver a solution. Some examples of transcendence within the storage management arena are automated provisioning, automated chargeback based on service-level agreements, policy-based management of issues within the storage environment; integrated application-based storage solutions or automated data protection. To get to this level of transcendence requires levels of management to be in place below the upper levels.

To start with, let's look at the lowest level, or physical level, as I like to call it. The physical level deals with all of the physical devices and discovering them and their associated properties. This process takes an inventory of every storage device (local disk, tape device, array, JBOD and so on), switch (IP for NAS and iSCSI or FC), the hosts with NICs or HBAs, applications installed and as much other physical information as possible. The more information that you have about an environment, the better able one is to make an intelligent decision. We will get to this later in the discussion or maybe in tomorrow's entry.

On the next layer, the second layer in the stack, we look to discover the virtual or logical information that is presented. For example, the volumes within the storage pool, zoning and LUN masking information, RAID setups, host-based mirroring and replication and any other logical information that can be gathered.

On the next level there are a bunch of common services that include the ability to correlate the information received from the two lower levels. This information gives you or an application the ability to draw a graphical representation of the environment. I am not sure who said it, but "A picture is truly worth a thousand words." With the picture, one is able to see the relationships between devices and add information that is user-defined, such as account codes, physical location and ownership information, which can enable higher-level functions, such as automated chargeback. Also, within this layer there are other functions such as event management, which deals with issues that are discovered in the environment, such as a path failure or device failure. Many reading this will start to think about Enterprise Management Systems (EMS) that were hot in the early 1990s. These systems provided all of these features on an enterprise scale, but they lacked input from the enterprise to the specific goals and processes within the organization. This was a critical part of their failure. They had some much to offer that to succeed in an implementation took a consulting team multiple years to implement the solution. So, these companies took their monolithic solutions and carved them up into packaged products that we know today.

To be continued Monday…
Posted by Brett Cooper Understanding products and solutions by using an organic principal to explain them
23 JUN 2005 22:39 EDT (02:39, GMT)
Maslow's Hierarchy of Needs has been used to explain everything from how transcendence occurs to the next generation of storage management technologies. I thought that it may be useful to utilize this principle to discuss products and how they become solutions. Many vendor Web sites today have two sections: products and solutions. "What's the difference," you may ask? I would say that a product is a single thing that a vendor makes, while a solution combines a product from the vendor with another product or partnership to create a solution to an enterprise challenge.

Understanding Maslow's Hierarchy of Needs

Simply put, the hierarchy above is a layered approach that states that in order to move from the lowest level of the hierarchy to a higher level, the needs in the lower level must first be met. For example, once an individual is born, they must establish the physiological needs and satisfy their hunger and thirst before they can move to safety needs and so on. In my opinion there are a variety of things missing from the above model, specifically that a couple (that we shall call the parents) must have been able to reach a higher level in order to procreate and create a new being in order for this new being to be able to begin their journey through the levels within the hierarchy. This will come up again in a moment as I explain how we can utilize this model to apply to storage management technologies….Hold onto your mouse and read on….

This layered or level-based approach is similar to the many storage management technologies that are available in the market today. First, we need to understand that many storage management technologies come from the creation of standards, which is another process in itself, which I will discuss in another blog entry later this week or early next, then a company works internally and with partners to develop their product and bring it to market with alpha and beta releases. In my opinion the process of bringing the product to market turns the hierarchy upside down until the product is delivered, at which point there is a layered approach to building higher-level functions that customers are looking for to manage their environments.

Tomorrow, look for the next installation of Maslow's Hierarchy of Needs for storage management, where we will look through the layered approach.
Posted by Brett Cooper How do you get things done? It's in the process…
22 JUN 2005 22:40 EDT (02:40, GMT)
Building on my entry from yesterday; where we discussed the first of the three Ps, the people, I wanted to move to the process. Many enterprises require processes for every part of their information technology solutions to be documented -- from provisioning new hosts within the environment to failing over to another site in the case of a disaster. Each day new processes come into play to correct issues as they arise or deliver new services to customers, both internal and external.

The key in understanding processes is to document them. I like to use the example of a military unit crossing a river. The first time a unit came to a river and said, "We need to get over that river," they did it using the best way that the management team could think of working with one of their engineers. The next time they came to a river they wanted to cross, they did so using the knowledge and process that they had learned from the units before them that were compiled into the Standard Operating Procedure, or SOP, for crossing a river. This process could then be updated with the information from the units that had crossed other rivers, and, as a whole, the military gained knowledge and was able to more efficiently and quickly get over the river and deliver the required resources to their troops at the front.

Think of this as a storage environment in which a storage administrator needs to deliver storage to a particular application on a host within the infrastructure. There is a standard process in place to deliver the required storage to the specific host. This process includes forms and processes that are specific to the environment. Moving beyond the storage processes, there are business processes that can effect the storage environment. Gaining alignment between the business processes and goals of the organization can help a storage administrator deliver a better level of service and expand an admin's insight into the overall environment. Further, if the processes are documented when individuals move around the organization, it is easier to keep the organization running through transitions.

What do you think of your processes in your storage environment? Are they documented? Tomorrow I'll take a look at the final P of the three Ps: the products.
Posted by Brett Cooper Creating solutions to information technology challenges takes people
21 JUN 2005 20:06 EDT (00:06, GMT)
With all the hype around automating solutions to the enterprise challenges of today, it should not be a surprise to anyone in the industry that consolidation is occurring. Some technologies can provide automated solutions to challenges in the industry, like thin provisioning and virtualization of resources from the storage to the host operating system platform. However, the simple challenges of provisioning storage within an enterprise still require a knowledgeable storage administrator who will carve out the right space on the right system, or systems, to deliver the required environment to meet the needs of the application or user -- performance, scalability, reliability or availability. Some of these solutions can take months to roll out within an enterprise, as they require integration between groups. Building the bridges between the storage group and the other functional areas within the enterprise takes patience and dedication as each group may report to the same CIO at the top of an organization. The CIO and the rest of the CXOs have the same goal: to create value for their shareholders, which translated to an administrator or manager in an organization, usually means to do more with less and execute flawlessly.

In time, I can see automation technologies handling some of the simpler tasks in an enterprise; but even the simpler tasks using the best-of-breed tools will require rules to operate efficiently and correctly. The relationship between rules and tools can be seen everywhere within the IT organization. In fact, I believe that EMC's purchase of System Management ARTS (SMARTS) is a way to automate the responses of the storage management environment to conditions that exist within the enterprise.

These rules can be tuned to meet the needs of a specific enterprise, which is where the people come into play. Each enterprise has similar but different goals and ways of doing business, which we can call the process. Also, the enterprise has different products that provide the platforms for business.

Aligning all three of these areas with the value of the information and then prescribing rules to the way the environment is supposed to work and what to do if it is not working correctly requires human involvement. One of the strongest calls for administrator involvement is in data protection, or backup and replication space. I believe it was Gartner that recently published that 50% to 80% of all backup jobs within the enterprise fail. Do you think this is true? If so, why? My guess is that it has to do with the medium of tape that the backup is being performed on as a start. (Did I replace the cleaning cartridge or rotate the media? Oh, dang it! That backup failed.) Add to that the complexity of the movement of the information from disk to tape. (I missed the backup window again!)

Thus the big play for disk-to-disk (D2D) backup in the industry today. Move the information from disk to a secondary tier of storage that costs less and provides better recoverability. Plus an organization can easily offsite a tape or replicate from site A to site B from my secondary system, never impacting my primary storage. All goodness, but again, back to needing the administrator to set up all of these tasks and then keep them going.

I remember a story that I had heard in college about the automobile manufacturers all moving to robotics in their manufacturing plants and saving hundreds of thousands of dollars in personnel payroll and benefits only to find that each robot costs millions of dollars and requires a well-paid administrator to keep the robot going, whereas hiring several skilled human beings costs the same as the robot administrator and the robot itself.

Why am I focusing on this topic in the blog today? I am interested in the balance between humanity and technology. The storage management technologies that are available in the market today are compelling, but they will require an administrator and tuning to fit the goals of the environment to return the maximum value. I also think that we all need to be on the lookout for continued consolidation within IT in the same way as it has appeared within the departmental organizations. With the move to a more application-focused storage environment, be on the lookout for application administrators to want to have a hand in the setup and management of the storage environment that supports their application. The application administrator already owns the application environment from a host perspective, and I would be willing to bet that in the near future there will be a merging of application and storage administrator roles and goals. Let me know if you agree with my thought or have another point of view.
Posted by Brett Cooper Storage management for the rest of us
20 JUN 2005 13:39 EDT (17:39, GMT)
Welcome to my first entry in my storage management blog! For the next two weeks I will be sharing some of my thoughts on the world of storage, and hopefully we can share thoughts and experiences and learn something in the process.

Having written for SearchStorage.com for some time, I thought it would be best to help you understand my point of view on the industry. We are all responsible for storage in one way or another, whether it is the local hard disk drive in our notebook computer or the RAID array that is storing the company's financial information, the process involved with managing storage can be as complex or simple as required by the goals of the information or data being stored. What is important to understand is how to determine the value of the information and then prescribe the right people, process and products (or, as I call them, the three Ps) to manage the storage environment. The value of the information can be determined by understanding the impact of the information on the business or person using it. Is the information critical to the function of the business? If so, then you may need to provide a solution that is not only highly available, but easily recovered in the case of an accidental deletion or virus or malware attack.

When I added the title to this entry, I thought that I would quick off a discussion and ask you what you think the definition of "storage management" is. My hope is that by the end of the next two weeks we can create a definition and some working roles and rules for our environments. How does that sound as a goal?
Posted by Brett Cooper