NOTE: My original article contained embedded calculation errors that significantly distorted the end results. These problems have since been corrected. I apologize to anyone who was accidentally misled by this information, and sincerely thank those diligent readers who brought the issues to my attention.
Some issues seem so obvious they’re hardly worth considering. Everyone knows that Solid State Drives (SSD) are more energy-efficient than spinning disk. They don’t employ rotating platters, electro-mechanical motors and mechanical head movement for data storage, so they must consume less power – right? However, everyone also knows the cost of SSD is so outrageous that they can only be deployed for super-critical high performance applications. But does the reputation of having exorbitant prices still apply?
While these considerations may seem intuitive, they are not entirely accurate. Comparing the Total Cost of Ownership (TCO) for traditional electro-mechanical disks vs. Solid State Disks provides a clearer picture of the comparative costs of each technology.
- For accuracy, this analysis compares the purchase price (CAPEX) and power consumption (OPEX) of only the disk drives, and does not include the expense of entire storage arrays, rack space, cooling equipment, etc.
- It uses the drive’s current “street price” for comparison. Individual vendor pricing may be significantly different, but the ratio between disk and SSD cost should remain fairly constant.
- The dollar amounts shown on the graph represent a 5-year operational lifecycle, which is fairly typical for production storage equipment.
- Energy consumption for cooling has also been included in the cost estimate, since it requires roughly the amount of energy as the drives consume to maintain them in an operational state.
- 100 TB of storage capacity was arbitrarily selected to illustrate the effect of cost on a typical mid-sized SAN storage array.
The following graph illustrates the combined purchase price, plus energy consumption costs for several popular electro-mechanical and Solid State Devices.
From the above comparison, several conclusions can be drawn:
SSDs are Still Expensive – Solid State Drives remain an expensive alternative storage medium, but the price differential between SSD and electro-mechanical drives is coming down. As of this writing there is only an x5 price difference between the 800GB SSD and the 600GB, 15K RPM drive. While this is still a significant gap, it is far less that the staggering x10 to x20 price differential seen 3-4 years ago.
SSDs are very “Green” – A comparison of the Watts consumed during a drive’s “typical operation” indicate that SSD consumes about 25% less energy than 10K RPM, 2.5-inch drives, and about 75% less power than 15K RPM, 3.5-inch disks. Given that a) each Watt used by the disk requires roughly 1 Watt of power for cooling to remove the heat that is produced, and b) the cost per Kwh continues to rise every year, this significant difference become a factor over a storage array-s 5-year lifecycle.
Extreme IOPS is a Bonus – Although more expensive, SSDs are capable of delivering from 10- to 20-times more I/O’s-per-second, potentially providing a dramatic increase in storage performance.
Electro-Mechanical Disks Cost Differential – There is a surprisingly small cost differential between 3.5 inch, 15K RPM drives and 2.5 inch 10K RPM drives. This may justify eliminating 10K disks altogether and deploying a less complex 2-tiered array using only 15K RPM disks and 7.2K disks.
Legacy 3.5 Inch Disks – Low capacity legacy storage devices (<146GB) in a 3-5-inch drive form-factor consume too much energy to be practical in a modern, energy-efficient data center (this includes server internal disks). Any legacy disk drive smaller than 300 GB should be retired.
SATA/NL-SAS Disks are Inexpensive – This simply re-affirms what’s already known about SATA/NL-SAS disks. They are specifically designed to be inexpensive, modest performance devices capable of storing vast amounts of low-demand content on-line.
The incursion of Solid State Disks into the industry’s storage mainstream will have interesting ramifications not only for the current SAN/NAS arrays, but also may impact a diverse set of technologies that have been designed to tolerate the limitations of an electro-mechanical storage world. As they say, “It’s a brave new world”.
Widespread deployment of SSD will have a dramatic impact on the storage technology itself. If SSDs can be implemented in a cost-effective fashion, why would anyone need an expensive and complex automated tiering system to decrement data across multiple layers of disk? Because of its speed, will our current efforts to reduce RAID rebuild times still be necessary? If I/O bottlenecks are eliminated at the disk drive, what impact will it have on array controllers, data fabric, and HBAs/NICs residing upstream of the arrays?
While it is disappointing to find SSD technology still commands a healthy premium over electro-mechanical drives, don’t expect that to remain the case forever. As the technology matures prices will decline when user acceptance grows and production volumes increase. Don’t be surprised to see SSD technology eventually eliminate the mechanical disk’s 40-year dominance over the computer industry.
For those of you interested in examining the comparison calculations, I’ve included the following spreadsheet excerpts contain detailed information used to create the graph.
Blade servers, virtualization, solid state disks, and 16Gbps fibre channel – it’s challenging to keep up with today’s advanced technology. The complexity and sophistication of emerging products can be dizzying. In most cases we’ve learned how to cope with these changes, but there are a few areas where we still cling to vestiges of the past. One of these relics of past decades is the impenetrable, monolithic data center.
The data center traces its roots back to the mainframe, when all computing resources were housed in a single, highly specialized facility designed specifically to support processing operations. Since there was little or no effort to classify data, these bastions of data processing were over-designed to ensure the most critical requirements were supported. This model was well-suited for mainframes and centralized computing, but it falls well short of meeting the needs of our modern IT environments.
Traditional data center facilities provide a one-size-fits-all solution. At an average $700 to $1500 per square foot, they are expensive to build. They lack the scalability and flexibility to respond to dynamic market changes and shifts in technology. Since these require massive investments of capital, they must be built not only to contain today’s IT equipment, but also satisfy growth requirements for 25-years or more. The end result is a tremendous waste of capacity, corporate funds tied up for decades, making assumptions about the direction and needs of future IT technology, the build-out of a one-size-fits-all facility, and a price tag that makes disaster recovery redundancy well beyond the reach of most companies.
An excellent solution to this problem is already a proven technology – the Portable Modular Data Center. These are typically self-contained data center modules that contain a comprehensive set of power, cooling, security, and internal infrastructure to support a dozen or more equipment racks per module with up to 30kW of power per rack. These units are relatively inexpensive, highly scalable, simple to deploy, energy efficient (Green), and factory constructed to ensure consistent quality and reproducible technology. As modules, they can be deployed incrementally as requirements dictate, avoiding major one-time capital expenditures for facilities.
Their inherent modularity and scalability make them an excellent choice for incrementally building out finely-tuned disaster recovery facilities. Here is an example of how modular data centers can be leveraged to cost-effectively provide Disaster Recovery protection of an organization’s data assets.
- Mission Critical Operations (typically 10% to 15%)
These are applications and data that might severely cripple the organization if they were not available for any significant period of time.
Strategy – Deploy synchronous replication technology to maintain an up-to-date mirror image of the data that could be brought to operational status within a matter of minutes.
Solution – Deploy one or more Portable Module Data Center units within 30-miles (to minimize latency) and run synchronous replication between the primary data center and the modular facility. Since 20-30 miles of separation would protect from a local disaster, but not a region-wide event, it might be worthwhile to replicate asynchronously from the modular data center to some remote (out-of-region) location. A small amount of data might be lost in the event of a disaster (due to asynchronous delay), but processing could still be brought back on-line quickly with minimal loss of data and only a limited interruption to operations.
- Vital Operations (typically 20% to 25%)
These applications and data are very important to the organization, but an outage of several hours would not financially cripple the business.
Strategy – Deploy an asynchronous replication mechanism outside the region to ensure an almost-up-to-date copy of data is available for rapid recovery.
Solution – Deploy one or more Portable Module Data Center units anywhere in the country and run asynchronous replication between the primary data center and the remote modular facility. Since distance is not a limiting factor for asynchronous replication, the modular facility could be installed anywhere. This protects from disasters occurring not only locally, but within the region as well. A small amount of data might be lost in the event of a disaster (due to asynchronous delay), but applications and databases could still be recovered quickly with minimal loss of data and only a limited interruption to operations.
- Sensitive Operations (typically 20% to 30%)
These applications and data are important to the organization, but an outage of several days to one week would have only a negligible financial impact on the business.
Strategy – (same as above) Use the same asynchronous replication mechanism outside the region to ensure an almost-up-to-date copy of data is available for rapid recovery.
Solution – Add one or more Portable Module Data Center units to the above facility (as required) and run asynchronous replication between the primary data center and the remote modular facility.
- Non-Critical Operations (typically 40% or more)These applications and data are incidental to the organization and can be recovered when time is available. An outage of several weeks would have little impact on the business.
Strategy – (same as above) Use the same asynchronous replication mechanism outside the region to ensure an almost-up-to-date copy of data is available for rapid recovery.
Solution – Deploy one or more Portable Module Data Center units anywhere in the country and run asynchronous replication between the primary data center and a remote modular facility.
Note: Since non-critical applications and data tend to be passive, non-critical operations might also be a viable candidate for transitioning to an Infrastructure-as-a-Service (IaaS) provider.
Modular Data Centers are the obvious enabler for the above Disaster Recovery strategy. They allow you to deploy only the data center resource you need, when you need it. They are less expensive than either leased or build facilities, and can be scaled as required by the business.
It’s time for the IT industry to abandon their outdated concepts of what a data center should be and focus on what is needed by each class of data. The day of raised-floor mainframe “bunkers” has passed. It’s time to start managing data center resource deployment as carefully as we manage server and storage deployment. Portable Modular Data Centers allow you to implement efficient, cost-effective IT production facilities in a logical sequence, without breaking the bank in the process.
The world of IT is becoming remarkably complex, and companies grow increasingly reliant on outside knowledge and skills for assistance. But when you enter into really uncharted waters and need someone you can trust, who will you call? Unfortunately there are a lot of companies in the industry claiming to be technical experts for everything from “Big Data” to “Desktop Virtualization”. How do you identify the serious resources from the technical wannabe’s?
That is an interesting question. Every since the industry’s rush to identify and fix Y2K problems over a decade ago, the line between Consultant, Contractor, and Staff Augmentation has blurred. The recession of the past few years further masked the distinction between roles, since many laid-off IT employees simply re-branded themselves as “Independent Consultants” in an attempt to secure short-term project work.
So what are the differences between Staff Augmentation, Contractors, and Consultants? Let’s start with a definition. According to Wikipedia:
“Staff Augmentation is an outsourcing strategy which is used to staff a project and respond to business objectives. The technique consists of evaluating the existing staff and then determining which additional skills are required. One possible advantage of this approach is that it may leverage existing resources as well as utilize outsourced services and contract workers.”
“An Independent Contractor is a natural person, business, or corporation that provides goods or services to another entity under terms specified in a contract or within a verbal agreement. Unlike an employee, an independent contractor does not work regularly for an employer but works as and when required, during which time he or she may be subject to the Law of Agency. Independent contractors are usually paid on a freelance basis.”
“A Consultant (from Latin: consultare “to discuss”) is a professional who provides professional or expert advice in a particular area such as security (electronic or physical), management, accountancy, law (tax law, in particular), human resources, marketing (and public relations), finance, engineering, or any of many other specialized fields. A consultant is usually an expert or a professional in a specific field and has a wide knowledge of the subject matter.”
…Wikipedia Online Dictionary
Staff augmentation is based on the concept of a “faceless, replaceable skill” that is available for an entire category of labor (Administrator, Engineer, Programmer, Database Administrator, Web Designer, etc.). Since IT relies on a large labor pool of technical skills, these are relatively low priced roles. Since participants are required to have only prerequisite skills in their specialty and no other unique capabilities, they can be hired and released pretty much on demand. Rates are dictated by current market prices, and range from $35 – $95 per hour.
IT contractors are further up the scale in capabilities and value. They are typically companies that deliver a complete service or system to solve a clearly defined problem. This may be a particular operation, type of application, virtualized infrastructure, or network operation. In many instances it is delivered as a complete package, including hardware, software, utilities, installation, configuration, and testing. Contractor services may be purchased on a per-project or a time-and-materials basis and are consistent with similar projects. Bundled labor rates within a specified package or service are in the $125 to $185 per hour range.
At the top of the pyramid is IT consultant. This is a professional service offering highly developed skills and extensive experience in a specialized field. In addition to being a Subject Matter Expert for a particular technology or service, IT consultants typically have an extensive knowledge of related activities that include business operations, project management, associated technologies, industry best practices, quality assurance, security, and other operations. They are sought out by organizations for their comprehensive understanding of business-critical operations or other activity than can have industry-changing ramifications. Since these are highly specialized skills, they command rates from $225 to $450 per hour or more. Although consultants are expensive, they return value to the company that can far exceed their billable rate.
Clearly it’s in the client’s best interests to understand the differences and capabilities of each category. Unfortunately, these titles are frequently intermixed and tossed around somewhat indiscriminately by organizations. Unless due diligence is performed beforehand, occasionally some hapless company will think they landed a senior Consultant for $85 per hr. (plus expenses), when they actually contracted Staff Augmentation. This can quickly becomes the root cause of poor performance, lack-luster productivity, poor organization, missed objectives, and ultimately a failed project.
Technical personnel do not automatically become senior consultants just because that’s a label they’ve anointed themselves with. Buyer beware! Engaging the proper skill-set can either be a game-changer, or a “boat anchor” for the project.
It’s hard to retire a perfectly good storage array. Budgets are tight, there’s a backlog of new projects in the queue, people are on vacation, and migration planning can be difficult. As long as there is not a compelling reason to take it out of service, it’is far easier to simply leave it alone and focus on more pressing issues.
While this may be the path of least resistance, it can come at a high price. There are a number of good reasons why upgrading storage arrays to modern technology may yield superior results and possibly save money too!
Capacity – When your aging disk array was installed several years ago, 300 GB, 10K RPM, FC disk drives were mainstream technology. It was amazing to realize you could squeeze up to 45 TB in a single 42U equipment rack! Times have changed. The same 10K RPM DISK drive has tripled in capacity, providing 900 GB in the same 3.5 inch disk drive “footprint”. It’s now possible to get 135 TB (a 300% capacity increase) into the same equipment rack configuration. Since data center rack space currently costs around $3000 per month, that upgrade alone will dramatically increase capacity without incurring any increase in floor-space cost.
Density – Previous generation arrays packaged from (12) to (15) 3.5 inch FC or SATA disk drives into a single rack-mountable 4U array. Modern disk arrays support from (16) 3.5 inch disks per 3U tray, to (25) 2.5 inch disks in a 2U tray. Special ultra-high density configurations may house up to (60) FC, SAS, or SATA DISK drives in a 4U enclosure. As above, increasing storage density within an equipment rack significantly increases capacity while requiring no additional data center floor-space.
Energy Efficiency – Since the EPA’s IT energy efficiency study in 2007 (Report to Congress on Server and Data Center Energy Efficiency, Public Law 109-431), IT manufacturers have increased efforts to improve the energy efficiency of their products. This has resulted in disk drives that consume from 25% to 33% less energy, and storage array controllers lowering power consumption by up to 30%. That has had a significant impact on energy costs, including not only the power to run the equipment, but also power to operate the cooling systems needed to purge residual heat from the environment.
Controller Performance – Storage array controllers are little more than specialized servers designed specifically to manage such functions as I/O ports, disk mapping, RAID and cache operations, and execution of array-centric internal applications (such as thin provisioning and snapshots). Like any other server, storage controllers have benefited from advances in technology over the past few years. The current generation of disk arrays contain storage controllers with from 3 to 5 times the processing power of their predecessors.
Driver Compatibility – As newer technologies emerge, they tend to focus on developing software compatibility with the most recently released products and systems on the market. With the passage of time, it becomes less likely for storage arrays to be supported by the latest and greatest technology on the market. This may not impact daily operations, but it creates challenges when a need arises to integrate aging arrays with state-of-the-art systems.
Reliability – Common wisdom used to be that disk failure characteristics could be accurately represented by a ”bathtub graph”. The theory was the potential for failure was high when a disk was new. It then flattened out at a low probability throughout the disk’s useful life, then took a sharp turn upswing as it approached end-of-life. This model implied that extending disk service life had no detrimental effects until it approached end-of-life for the disks.
However over the past decade, detailed studies by Google and other large organizations with massive disk farms have proven the “bathtub graph” model incorrect. Actual failure rates in the field indicate the probability of a disk failure increases by 10% – 20% for every year the disk is in service. It clearly shows the probability of failure increases in a linear fashion over the disk’s service life. Extending disk service-life greatly increases the risk for disk failure.
Service Contracts –Many popular storage arrays are covered by standard three-year warranties. This creates a dilemma, since the useful service life of most storage equipment is considered to be either four or five years. When the original warranty expires, companies must decide whether to extend the existing support contract (at a significantly higher cost), or transitioning to a time & materials basis for support (which can result in some very costly repairs).
Budgetary Impact – For equipment like disk arrays, it is far too easy to fixate on replacement costs (CAPEX), and ignore the ongoing cost of operational expenses (OPEX). This may avoid large upfront expenditures, but it slowly bleeds the IT budget to death by having to maintain increasingly inefficient, fault-prone, and power hungry equipment.
The solution is to establish a program of rolling equipment replenishment on a four- or five-year cycle. By regularly upgrading 20% to 25% of all systems each year, the IT budget is more manageable, equipment failures are controlled, and technical obsolescence remains in check.
Getting rid of familiar things can be difficult. But unlike your favorite slippers, the LazyBoy recliner, or your special coffee cup, keeping outdated storage arrays in service well beyond their prime can cost your organization plenty.
It is somewhat surprising just how many skilled IT specialists still shy away from eliminating traditional internal boot disks with a Boot-from-SAN process. I realize old habits die hard and there’s something reassuring about having the O/S find the default boot-block without needing human intervention. However the price organizations pay for this convenience is not justifiable. It simply adds waste, complexity, and unnecessary expense to their computing environment.
Traditionally servers have relied on internal disk for initiating their boot-up processes. At start-up, the system BIOS executes a self-test, starts primitive services like the video output and basic I/O operations, then goes to a pre-defined disk block where the MBR (Master Boot Record) is located. For most systems, the Stage 1 Boot Loader resides on the first block of the default disk drive. The BIOS loads this data into system memory, which then continues to load Stage 2 Boot instructions and ultimately start the Operating System.
Due to the importance of the boot process and the common practice of loading the operating system on the same disk, two disks drives with a RAID1 (disk mirroring) configuration is commonly used to ensure high availability.
Ok, so far so good. Then what’s the problem?
The problem is the disks themselves. Unlike virtually every subsystem in the server, these are electro/mechanical devices with the following undesirable issues:
- Power & Cooling – Unlike other solid-state components, these devices take a disproportionately large amount of power to start and operate. A mirrored pair of 300GB, 15K RPM disks will consume around .25 amps of power and need 95.6 BTUs for cooling. Each system with internal disk has its own miniature “space heater” that aggravates efforts to keep sensitive solid state components cool.
- Physical Space – Each 3.5 inch drive is 1” x 4.0” x 5.76” (or 23.04 cubic inches) in size, so a mirrored pair of disks in a server represents an obstacle of 46.08 cubic inches that requires physical space, provisions for mounting, power connections, air flow routing, and vibration dampening to reduce fatigue on itself and other internal components.
- Under-utilized Capacity – As disk drive technology continues to advance, it becomes more economical to manufacture higher capacity disk drives than maintain an inventory of lower capacity disks. Therefore servers today are commonly shipped with 300GB or 450GB boot drives. The problem is that Windows Server 2008 (or similar) only needs < 100GB of space, so 66% of the disk’s capacity is wasted.
- Backup & Recovery – Initially everyone plans to keep only the O/S, patches and updates, log files, and related utilities on the boot disk. However, the local disk is far too convenient and eventually has other files “temporarily” put on it as well. Unfortunately some companies don’t include boot disks in their backup schedule, and risk losing valuable content if both disks are corrupted. (Note: RAID1 protects data from individual disk failures but not corruption.)
Boot-from-SAN does not involve a PXE or tftp boot over the network. It is an HBA BIOS setting that allows SAN disk to be recognized very early in the boot process as a valid boot device, then points the server to that location for the Stage 1 Boot Loader code. It eliminates any need for internal disk devices and moves the process to shared storage on the SAN. It also facilitates the rapid replacement of failed servers (all data and applications remain on the SAN), and is particularly useful for blade systems (where server “real-estate” is at a premium and optimal airflow is crucial).
The most common argument used against Boot-from-SAN is “what if the SAN is not available”. On the surface it sounds like a valid point, but what is the chance of that occurring with well-designed SAN storage? Why would that be any different than if the internal boot disk array failed to start? Even if the system started internally and the O/S loaded, how much work could a server do if it could not connect to the SAN? The consequences of any system failing to come up to an operational state are the same, regardless if it uses a Boot-from-SAN process or boots up from internal disks.
For a handful servers, this may not be a very big deal. However, when you consider the impact on a datacenter running thousands of servers the problem becomes obvious. For every thousand servers, Boot-from-SAN eliminates the expense of two thousand internal disks, 240 amps of current, the need for 655,300 BTUs of cooling, greatly simplifies equipment rack airflow, eliminates 200TB of inaccessible space, and measurably improves storage manageability and data backup protection.
Boot-from-SAN capability is built into most modern HBA BIOS’s and is supported by almost every operating system and storage array on the market. Implementing this valuable tool should measurably improve the efficiency of your data center operation.
I’m frequently surprised by the number of companies who haven’t transitioned to a tiered storage structure. All data is not created equal. While a powerful database may place extreme demand on storage, word processing documents do not.
As we move into a new world of “big data”, more emphasis needs to be focused on making good decisions about what class of disk this data should reside on. Although there are no universally accepted standards for storage tier designations, frequently the breakdown goes as follows:
Tier 0 – Solid state devices
Tier 1 – 15K RPM SAS or FC Disks
Tier 2 – 10K RPM SAS or FC Disks
Tier 3 – 7200 or 5400 RPM SATA (a.k.a. – NL-SAS) Disks
So why is a tiering strategy important for large quantities of storage? Let’s take a look at similar storage models for 1 petabyte of data:
The difference in disk drive expense alone is over $225,000 or around 30% of the equipment purchase price. In addition there other issues to consider.
- Reduces the Initial purchase price by 25% or more
- Improving energy efficiency by 25% – 35% lowers operational cost and cooling requirements
- Substantial savings from reduced data center floorspace requirements
- Increased overall performance for all applications and databases
- Greater scalability and flexibility for matching storage requirements to business growth patterns
- Provides additional resources for performance improvements (an increased number of ports, cache, controller power, etc.)
- A high degree of modularity facilitates better avoidance of technical obsolescence
- May moderate the demand for technical staff necessary to manage continual storage growth
- Requires automated, policy-based data migration software to operate efficiently.
- Should employ enterprise-class frames for Tiers 0/1 and midrange arrays for Tiers 2/3
- Incurs approximately a 15% cost premium for enterprise-class storage to support Tier 0/1 disks
- Implements a more complex storage architecture that requires good planning and design
- Needs at least a rudimentary data classification effort for maximum effectiveness
So does the end justify the effort? That is for each company to decide. If data storage growth is fairly stagnant, then it may be questionable whether the additional effort and expense is worth it. However if you are staggering under a 30% – 50% CAGR storage growth rate like most companies, the cost reduction, increased scalability, and performance improvements achieved may well justify the effort.