The Software RAID Inflection Point

In June, 2001 something amazing happened in the IT world: Intel released the Tualatin based Pentium IIIS 1.0 GHz processor. This was one of the first few Intel processors (IA32 architecture) to have crossed the 1 GHz clock barrier and the first of any significance. It was also special in that it had dual processor support and a double sized cache compared to its Coppermine based forerunners or it’s non-“S” Tualatin successor (that followed just one month behind.) The PIIIS system boards were insanely popular in their era and formed the backbone of high performance commodity servers, such as Proliant and PowerEdge, in 2001 and for the next few years culminating in the Pentium IIIS 1.4GHz dual processor systems that were so important that they resulted in kicking off the now famous HP Proliant “G” naming convention. The Pentium III boxes were “G1”.

What does any of this have to do with RAID? Well, we need to step back and look at where RAID was up until May, 2001. From the 1990s and up to May, 2001 hardware RAID was the standard for the IA32 server world which mainly included systems like Novell Netware, Windows NT 4, Windows 2000 and some Linux. Software RAID did exist for some of these systems (not Netware) but servers were always struggling for CPU and memory resources and expending these precious resources on RAID functions was costly and would cause applications to compete with RAID for access and the systems would often choke on the conflict. Hardware RAID solved this by adding dedicated CPU and RAM just for these functions.

RAID in the late 1990s and early 2000s was also very highly based around RAID 5 and to a lesser degree, RAID 6, parity striping because disks were tiny and extremely expensive for capacity and squeezing maximum capacity out of the available disks was of utmost priority and risks like URE were so trivial due to the small capacity sizes that parity RAID was very reliable, all things considered. The factors were completely different than they would be by 2009. In 2001, it was still common to see 2.1GB, 4.3GB and 9GB hard drives in enterprise servers!

Because parity RAID was the order of the day, and many drives were typically used on each server, RAID had more CPU overhead on average in 2000 than it did in 2010! So the impact of RAID on system resources was very significant.

And that is the background. But in June, 2001 suddenly the people who had been buying very low powered IA32 systems had access to the Tualatin Pentium IIIS processors with greatly improved clock speeds, efficient dual processor support and double sized on chip caches that presented an astounding leap in system performance literally over night. With all this new power and no corresponding change in software demands systems that traditionally were starved for CPU and RAM suddenly had more than they knew how to use, especially as additional threads were available and most applications of the time were single threaded.

The system CPUs, even in the Pentium III era, were dramatically more powerful than the small CPUs, which were often entry level PowerPC or MIPS chips, on the hardware RAID controllers and the available system memory was often much larger than the hardware RAM caches and investing in extra system memory was often far more effective and generally advantages so with the availability of free capacity on the main system RAID functions could, on average be moved from the hardware RAID cards to the central system and gain performance, even while giving up the additional CPU and RAM of the hardware RAID cards. This was not true on overloaded systems, those starved for resources and was more relevant for parity RAID systems with RAID 6 benefiting the most and non-parity systems like RAID 1 and 0 benefiting the least.

But June, 2001 was the famous inflection point – before that date the average IA32 system was faster when using hardware RAID. And after June, 2001 new systems purchased would on average be faster with software RAID. With each passing year the advantages have leaned more and more towards software RAID with the abundance of underutilized core CPUs and idle threads and spare RAM exploding with the only advantage towards hardware RAID being the drop in parity RAID usage as mirrored RAID took over as the standard as disk sizes increased dramatically while capacity costs dropped.

Today is has been more than fifteen years since the notion that hardware RAID would be faster has been retired. The belief lingers on due primarily to the odd “Class of 1998” effect. But this has long been a myth repeated improperly by those that did not take the time to understand the original source material. Hardware RAID continues to have benefits, but performance has not been one of them for the majority of the time that we’ve had RAID and is not expected to ever rise again.

Legitimate University Programs Are Not Certification Training

The university educational process is one that is meant to broaden the mind, increase exposure to different areas, teach students to think outside of the box, encourage exploration, develop soft skills, and to make students better prepared to tackle more learning such as moving on to trade skills needed for specific fields.  The university program, however, is not meant to provide trade skills themselves (the skills used in specific trades), that is the role of a trade school.   Students leaving universities with degrees are intended to not be employable due to specific skill sets learned at college, but to be well prepared to learn on the job or move on to additional education for a specific job.

In the last two decades, led primarily by for profit schools looking to make money quickly without regards to the integrity of the university system, there has been a movement, especially in the United States, for trade schools to get accredited (an extremely low bar requirement that has no useful standing outside of legal qualifications for educational minimums and should never be see as a mark of quality) and sell trade degrees as if they were traditional university degrees.  This has been especially prevalent in IT fields where certifications are broadly known and desired, acquiring properly skilled educational staff is expensive and essentially impossible to do at the scale necessary to run a full program, degree areas are easily misunderstood by those entering their college years and where the personality traits most common to people going into the field sadly makes those people easy prey for collegiate marketing drives.  The promise of easy classes, double dipping (getting the certs you need anyway then getting a bonus degree for the effort) and the suggestion that by having a degree and certs all at once will open doors and magically provide career options that pay loads of money triggers an emotional response that makes potential students less able to make rational financial and education decisions, additionally.  It’s a predatory market, not an altruistic one.

Certificates play a fundamentally different role than a university education does.  Unlike universities, certification is about testing very specific skills, often isolated by product or vendor, things that should never appear in any university program.  Certification may be broad (and closer to collegiate work) in certs like the CompTIA Network+ which tests a broad range of basic networking knowledge and nothing specific to a vendor or product, but is still overly specific to a single networking technology or group of technologies to be truly appropriate for a university, but is, at the very least, leaning in that direction.  But more common certifications such as Microsoft MCSE, Cisco’s CCNA, CompTIA’s Linux+ or A+ are all overly product and vendor specific, far too “which button do I press” and far too little “what does the underlying concepts mean” for collegiate work.

Certifications are trade related and a great addition to university studies.  University work should prepare the student for broad thinking, critical thinking, problem solving and core skills like language, maths and learning.  Then applying that core knowledge to certifications should make achieving certifications easier and meaningful.  University should show a background in soft skills and broadness, while certifications should show trade skills and specific task capabilities.

Warning signs that a university is behaving improperly would include, in regards to this area of concern, overly specific programs that sound as if they are aimed at technologies like a degree in “Cisco Networking” or “Microsoft Systems”, if certifications are achieved during the university experience (double dipping – giving out a degree simply for having gotten certs) or if the program leans towards an indication of preparing someone “for the job” or expected to “get the student a great job upon completion” or is expected to “increase salary”.  These are not goals of proper university programs.

Critically evaluating any educational program is very important as educational investments are some of the largest that we make in our lives, both monetarily and in terms of our time commitments.  Ensuring that the programs are legitimate, valuable, meet both our own goals and proper goals, will be seen as appropriate by those that will see them in the future (such as hiring managers) are very important.  There are many aspects over which we must evaluate the university experience, this is only one but it is one that is a newer problem, suddenly very prevalent and one that specifically targets IT and technical hopefuls so requires extra diligence in our industry.

 

Understanding Technical Debt

From Wikipedia: “Technical debt (also known as design debt or code debt) is “a concept in programming that reflects the extra development work that arises when code that is easy to implement in the short run is used instead of applying the best overall solution”.

Technical debt can be compared to monetary debt. If technical debt is not repaid, it can accumulate ‘interest’, making it harder to implement changes later on. Unaddressed technical debt increases software entropy. Technical debt is not necessarily a bad thing, and sometimes (e.g., as a proof-of-concept) technical debt is required to move projects forward. On the other hand, some experts claim that the “technical debt” metaphor tends to minimize the impact, which results in insufficient prioritization of the necessary work to correct it.”

The concept of technical debt comes from the software engineering world, but it applies to the world of IT and business infrastructure just as much. Like software engineering, we design our systems and our networks, and taking shortcuts in our designs, which includes working with less than ideal designs, incorporating existing hardware and other bad design practices produce technical debt.  One of the more significant forms of this comes from investing in the “past” rather than in the “future” and is quite often triggered through the sunk cost fallacy (a.k.a. throwing good money after bad.)

It is easy to see this happening in businesses every day.  New plans are made for the future, but before they are implemented investments are made in making an old system design continue working, work better, expand or whatever.  This investment then either turns into a nearly immediate financial loss or, more often, becomes incentive to not invest in the future designs as quickly, as thoroughly or possible, at all.  The investment in the past can become crippling in the worst cases.

This happens in numerous ways and is generally unintentional.  Often investments are needed to keep an existing system running properly and, under normal conditions, would simply be made.  But in a situation where there is a future change that is needed or potentially planned this investment can be problematic.  Better cost analysis and triage planning can remedy this, in many cases, though.

In a non-technical example, imagine owning an older car that has served well but is due for retirement in three months.  In three months you plan to invest in a new car because the old one is no longer cost effective due to continuous maintenance needs, lower efficiency and so forth.  But before your three month plan to buy a new car comes around, the old car suffers a minor failure and now requires a significant investment to keep it running.  Putting money into the old car would be an new investment in the technical debt.  Rather than spending a large amount of money to make an old car run for a few months, moving up the time table to buy the new one is obviously drastically more financially sound.  With cars, we see this easily (in most cases.)  We save money, potentially a lot of it, by quickly buying a new car.  If we were to invest heavily in the old one, we either lose that investment in a few months or we risk changes our solid financial planning for the purchase of a new car that was already made.  Both cases are bad financially.

IT works the same way.  Spending a large sum of money to maintain an old email system six months before a planned migration to a hosted email system would likely be very foolish.  The investment is either lost nearly immediately when the old system is decommissioned or it undermines our good planning processes and leads us to not migrate as planned and do a sub-par job for our businesses because we allowed technical debt to drive our decision making rather than proper planning.

Often a poor triage operation or improper authority to triage players can be the factor that causes emergency technical debt investments rather than rapid future looking investments.  This is only one area where major improvements may address issues, but it is a major one.  This can also be mitigated, in some cases, through “what if” planning to have investment plans in place contingent on common or expected emergencies that might arise, which may be as simple as capacity expansion needs due to growth that happen before systems planning comes into play.

Another great example of common technical debt is server storage capacity expansion.  This is a scenario that I see with some frequency and demonstrates technical debt well.  It is common for a company to purchase servers that lack large internal storage capacity.  Either immediately or sometime down the road more capacity is needed.  If this happens immediately we can see that the server purchased was a form of technical debt in improper design and obviously represents a flaw in the planning and purchasing process.

But a more common example is needing to expand storage two or three years after a server has been purchased.  Common expansion choices include adding an external storage array to attach to the server or modifying the server to accept more local storage.  Both of these approaches tend to be large investments in an already old server, a server that is easily forty percent or higher through its useful lifespan.  In many cases the same or only slightly higher investment in a completely new server can result in new hardware, faster CPUs, more RAM, the storage needed, purpose designed and built, aligned and refreshed support lifespan, smaller datacenter footprint, lower power consumption, newer technologies and features, better vendor relationships and more all while retaining the original server to reuse, retire or resell.  One way spends money supporting the past, the other often can spend comparable money on the future.

Technical debt is a crippling factor for many businesses.  It increases the cost of IT, sometimes significantly, and can lead to high levels of risk through a lack of planning and most spending being emergency based.

 

New Hyperconvergence, Old Storage

We all dream of the day that we get to build a new infrastructure from the ground up without any existing technical debt to hold us back.  A greenfield deployment where we pick what is best, roll it out fresh and enjoy.  But most of us live in the real world where that is not very realistic and what we actually face is a world where we have to plan for the future but work with what we already have, as well.

Making do with what we have is nearly an inevitable fact of life in IT and when approaching storage when moving from an existing architecture to hyperconvergence things will be no different.  In a great many cases we will be facing a situation where an existing investment in storage will be in place that we do not want to simply discard but does not necessarily fit neatly into our vision of a hyperconverged future.

There are obvious options to consider, of course, such as returning leased gear, retiring older equipment or selling still useful equipment outright.  These are viable options and should be considered.  Eliminating old gear or equipment that does not fit well into the current plans can be beneficial as we can simplify our networks, reduce power consumption and possible even recoup some degree of our investments.

In reality, however, these options are rarely financially viable and we need to make more productive use of our existing technology investments.  What options are available to us, of course, depend on a range of factors.  But we will look at some examples of how common storage devices can be re-purposed in a new hyperconverged-based system in order to maintain their utility either until they are ready to retire or even, in some cases, indefinitely.

The easiest re-purposing of existing storage, and this applies equally to both NAS and SAN in most cases, is to designate them as backup or archival targets.  Traditional NAS and SAN devices are excellent backup hardware and are generally usable by nearly any backup mechanism, regardless of approach or vendor.  And because they are generic backup targets if a mixture of backup mechanisms are used, such as agent based, agentless and custom scripts, these can all work to the same target.  Backups so rarely get the attention and investment that they deserve that this is not just the easiest but often the most valuable use of pre-existing storage infrastructure.

Of course anything that is appropriate for backups can also be used for archival storage.  Archival needs are generally less needed (only a percentage of firms need archival storage while all need backups) and are of lower priority, so this is more of an edge reuse case, but still one to consider, especially for organizations that may be working to re-purpose a large number of possibly disparate storage devices.  However it is worth noting that moving to hyperconvergence does tend to “flatten” the compute and storage space in a way that may easily introduce a value to lower performance, lower priority archival storage that may not have existed or existed so obviously prior to the rearchitecting of the environment.

NAS has the unique advantageous use cases of being usable as general purpose network storage, especially for home directories of end users.  NAS storage can be used in so many places on the network, it is very easy to continue to use, after moving core architectures.  The most popular case is for users’ own storage needs with the NAS connected directly to end user devices which allows for storage capacity, performance and network traffic to be offloaded from the converged infrastructure to the NAS.    It would actually be very rare to remove a NAS from a hyperconverged network as its potential utility is so high and apparent.

Both SAN and NAS have the potential to be attached directly to the virtual machines running on top of a hyperconverged infrastructure as well.  In this way they can continue to be utilized in a traditional manner until such time as they are no longer needed or appropriate.  While not often the recommended approach, attaching network storage to a VM directly, there are use cases for this and it allows systems to behave as they always have in a physical realm into the future.  This is especially useful for mapped drives and user directories via a NAS, much as we mentioned for end user devices, but the cases are certainly not limited to this.

A SAN can provide some much needed functionality in some cases for certain workloads that require shared block storage that otherwise is not available or exposed on a platform.  Workloads on a VM will use the SAN as they always have and not even be aware that they are virtualized or converged.  Of course we can also attach a SAN to a virtualized file server or NAS head running on our hyperconverged infrastructure if the tiering for that kind of workload is deemed appropriate as well.

Working with existing infrastructure when implementing a new one does present a challenge, but one that we can tackle with creativity and logical approach.  Storage is a nearly endless challenge and having existing storage to re-purpose may easily end up being exceptionally advantageous.