Archive for December, 2008

Never Fall A Sleep At Work After A Big Office Holiday Party!

Christmas Eve is upon us yet again, and I leaned back in my chair, put up my feet, balancing somewhat poorly, in my standard corporate chair, which is not designed for leisure repose. The boss via instance messaging, and all my other co-workers, few as they are now, that come into the office routinely, are long gone for the day, at last. I have a mug of tea heated to perfection, now at my elbow on my desk. I am content; another year of successful virtualization engineering and implementation is done. VMworld was a blast; to be sure, all things considered it has been a good year. So relaxed am I, proud of the achievements thus far this year, I feel the creep of sleep approach my consciousness… I start to think this is not the place to dream… but this thought is never finished as I slip into unconsciousness…

I sat in my favorite chair, in fact, the only chair in my bed chamber, the last bite of my warmed gruel, still favoring my tongue. The cold of the house barely avoided by the weak flames in the fireplace, I bemoaned the high price of cool. I hear the wind out side, move across the house structure, but I am warm, comfortable in my miserly surroundings. Unexpectedly, I felt a sudden chill, a phantasm like wave of cold. I fear for my life, but no rational evidence to do so is apparent, or so I thought. Faintly and then louder, I heard the clanking and jingling of metal objects linked by almost silent, fiber optic cable, the scratch or scrape of heavy objects across the hard oak floor. The objects, which appeared to be servers of various sizes and models, all dragged by a ghostly image of the contractor technical support person we, lost to the severe cold of the server vault all those years ago. As this person, of whom I fail to recall their given name, perished while struggling to rack hundreds of servers in just a couple of days. A cruel end for a poor soul that was just doing as instructed. Never mind the task was impossible for any mere mortal to do alone.

This poor unfortunate soul, this poltergeist condemned to travel the spiritual waste lands of the nether world in server support hell, approached me as though solid walls had no substance, having come through my bedroom door as if it was nothing more than smoke or mist. Again the noise of the looped together servers and other miscellaneous fiber fabric infrastructure, I now recognize, assaults my ears. I am speechless, the enjoyed gruel after taste in my mouth, now soured on my tongue. As the apparition spoke with haunting, hollow tones, words drawn out like they were painful to speak, and slurred in duress, as though each syllable was an extreme effort of will to be formed.

It, the thing unreal shadow, said “You will be visited by three wraiths of the virtual realm. They will come after the midnight hour but before dawn, each in its own time. Beware, beware, these wraiths will part the veil of reality. Be warned, be warned, that you heed their words of wisdom and advice, so you have been warned, so you should be aware. Fore ignoring their respective councils is doom assured, and thus doom earned.”

I squealed in fright, I jumped from my comfortable, if suspected quality of chair, and made flight to my bed, drawing the curtains of the bed frame around me, hiding my sight from the specter that voiced such dire prophetic whispers. No, I thought, no, this can not be real, it is not real, I am a man, a man of significant years. I can not and will not be boxed into a corner like timid child. I thought to myself, be gone, be gone you unholy spirit, I have no need of your contempt or compassion. But I stayed my tongue, voicing no sound, for I was, I admit, afraid. And as I thought these thoughts to my self, I heard again the noises I heard before, the rattle, clanking, and merciless scratching, of the technical resource ghost that was once human make its exit as it had its entrance. I thought anew, had this same technical position been not out-sourced under the circumstances then done, this poor entity might yet live still?

Some time passed, and I don’t recall the old clock on the fireplace mantel chiming at all, but when the chime toned thrice, I realized it was now three(3) in the morning or there about? For the old clock was wont to keep imperfect time. As poor in resources I felt I was, I had no desire to acquire a more accurate time piece. After all, approximate time, is good enough, only a fool desires the real time, all the time. A penny saved, is a penny earned, no? But my self congratulation for being frugal was interrupted, and I do not reference such as a computer processor would be. For a physical presence was near me, just out side the curtains of my bed I believed, nay, I felt there. In trembling movements I opened the curtains and there before me, was a wraith, as promised before. I yelled in spite of myself, and hid my direct view with a nearby pillow, saying “Go away oh wraith, leave me in peace. I deserve not this visitation.”

The wraith, in response did nothing more than motion with an appendage that appeared to be a very old and weathered hand, that I should come forth. No word or sound was made, nothing but an android or robotic like stiff jester with the hand. I could not resist the beckoning, though I hated being intrigued by this event. As I left the false security of my bed, I donned my robe and slippers, leaving my night hat a top my head, I also attired myself in my shabby old robe that was at hand at the end of the bed, where I had left it before sleep. Once I seemed prepared, the wraith placed that same aged hand against my forehead, and I lost all since of time and space, as I closed my eyes, for this was an uncomfortable happening. Movement of some type I knew was done, but the means of transportation and location selection were beyond my observation or ability to describe. It felt like I was falling, losing balance, but I never seemed to fall. After some brief time of this sensation, I sensed it was time to open my eyes. I did so. But what I say was both irrational and yet believable. Something I could some how identify as true.

Before me aligned in straight rows, some 100s, no 1000s of server racks, in long rows, stretched across a room longer and wider than I could determine with unassisted vision. Every make and model of server were represented, all manner of distributed server by vendors I recognized were in this unworldly presentation. I walked up to the nearest row of these racks, looking at the enclosed devices, and touched the keyboard nearest me. The corresponding flat-panel, be it small, and quite old in appearance did flicker into refractive light, the backlighting of screen still functional, a miracle given its apparent age. Before me was the simple console of a VMware Virtual Infrastructure (VI) ESX or it appeared to be such? It was hard to tell the specific version, since no reference to the version of ESX was obvious on the screen, this struck me odd, that I could see it was a VMware VI host before me, executing, but I could not focus on the specific version. All I gathered from the screen that this was old infrastructure; this entire location was in fact very old, the distant past, in a word, obsolete? The wraith, still not saying a single word, nor making any sound, motioned to all the racks in a wide sweeping movement, that imparted to my understanding that the entire room, as such, was all the same, VMware VI hosts in all its forms and variances was shown before me.

As I worked the keys on the keyboard before me, I realized in horror, that even though the host I connected to at random, was functional, no virtual instances were in existence. This realization allowed a cold chill to walk down my spine. “Was in fact, this entire expanse empty of all virtual reality? Were there not virtual instances in existence here?” I did not realize I had thought this idea aloud, but I had. Again, I asked, “Were there not virtual instances in existence here?” The wraith moved not the slightest, nor gave any indication that my question was acknowledged or understood. Still, I asked, pointing to the rack both to the left and right of where I stood saying, “Pray, tell me wraith, is this the future? Is VMware not but doomed to be abandoned?” Again, as before, the wraith gave no response. I yelled in frustration, “Wraith before me, speak!” Still this elicited no response.

After a few moments more, the wraith raised that same aged hand, a move I was now familiar with, and motioned for me to follow, I did this, taking one step then another with honest dread as to what I would see next. I was led to a cubical via that same falling sensation of transportation. This cube was long layered in dust and neglect, for it appeared, had not been used for some significant length of time. The trash bins were over flowing with papers and empty cans of Jolt cola. The desk was covered in corporate snail mail and other correspondence from various vendors, unsolicited marketing information to be sure, it all. An empty coffee cup placed at the edge of the desk, which read simple script, “We have done so much, for so long, we can now do anything with nothing, and management could still, care less!” It was clear this was a dead-end cube, which some unfortunate life form once used but was now long forgotten, long abandoned to a fate unknown. The wraith pointed to the name plate on the cube wall just at the entrance to same. Reading the name there, I fell to my knees, the strength of my legs abandoning me. Disregarding the pain of this maneuver, crying in a weak voice I said, “No, no. Please spirit, no.” I was in shock, and despaired at what I had just read. Saying with difficulty around my tears, “Spirit tell me this is not true, that this is only what could be?” But as before, the phantasm before me said nothing, making no jester or response in any way as recognition of my query. I felt confused, I felt lost.

The next instant or so it seemed to me, I awoke. I was sitting in my chair, in my cube. I found it odd that I recalled nothing about two (2) of the three (3) wraiths in my dream. Was I visited by all three or just one? No sooner did I think this thought, than a cold chill walked up my spine, where before it walked only down. The same chill? It changed directions? But that was in a dream, I told myself. Nothing but a dream, I told my self again. With practiced ease, I pulled my legs down to their proper business position, if only the boss knew how often I have my feet above knees with my keyboard placed across my hips as I sit in a relaxed pose, while at work. But the boss is 1000s of miles away, never yet to visit my physical location. So be it. With little thought to other needs of work, I started to disconnect my laptop, planning to stuff same into my portable bag. Thinking I never know when I might need to access the company network over the holidays. When my eye just catches one specific iconic flash in the bottom right corner… an icon that warns that I have new mail, yet unread. I freeze, I think, then say to no one but myself, “Ah, dang it, ignore it, it is Christmas day!” But truth be told, my personified view of life is closer to that of a cat, than of some other animal, whereas my cursorily often gets the better of me.

I click the icon, thinking, with luck it is something of no consequence, something that will be banished to the delete folder or little better. However, it is addressed to all my peers, my manager and other managers in the department, as well as me. It is from the boss of my boss, a communiqué from the top of my world, which makes me nervous. For as everyone knows this type of note, this time of year, is never good news, or never has been in the past. Regardless of my fears, and because I clicked the icon, the note opens to a larger window, which displays the message…

“Everyone, sorry for the late notice, but I wanted to get this to you all before the end of year. It has been decided, that starting the first of next month, we will begin the process of implementing Xen, and Hyper-V as our core virtualization platform, variable class of service, for virtual instances that leverage the Windows operating system. Furthermore, all Solaris virtual instances now run on VMware VI will be retired, and functionality migrated to Solaris LDOMs. All Linux virtual instances on VMware VI will be migrated or otherwise recreated as RHEL KVM based instances, in parallel, to the Solaris move.” As I read this, my knees start to shake; my hands take on a noticeable tremor. What? Has VMware VI not served us well? I read on, having trouble focusing on the meaning behind the words, “I am sure everyone saw this coming, so this should not be a surprise to anyone. Please understand the timeline for this is very aggressive, we need to get this done sooner than later. Right after the holidays, we will have a few meetings to work out the tactical objectives now that the strategic direction has been established. Remember, we have to get more done this coming year than ever before, faster than ever before, or else. Oh, I want to also say, of course, to everyone, Happy Holidays!” Is it a dream, did the wraith have it right? What is real? What is virtual? I am again scared…

Abruptly, I scream in pain, this hurts. No this really hurts… I am now, really, awake! I realize I have fallen out of my chair, spilling my almost empty cup of tea on myself in the process. What a horrible dream? Yes, a dream about a dream? As I set to rights my chair, and clean up the last of the spilt tea, I think to myself… Never fall a sleep at work after a big office holiday party… It is then, and only then, I notice that infernal inhuman soulless iconic flashing at bottom right side of the flat panel monitor on my desk. The unread electronic mail indicator is demanding immediate attention.

3 comments December 24th, 2008

Cloud Computing, Utility Computing, Bah Humbug!

Virtualization Critical Comparison – Chapter 03

Well, boys and girls, of which there should be more girls in the information technology industry right? Christmas is upon us, and the New Year will be here faster than you can say Jack Frost. Thanksgiving is gone, Halloween is long gone, so where are we? We are in that dead space between the old and new, when all planning is up in the air, resource planning, time planning, project planning are all but done, but could change over night this time of year, here to day, gone tomorrow, is a real situation this time of year. Wonderful, but what have we done to change the world, in the information technology industry as a whole? What will we do in 2009 that will shake the foundation of a strategic planning? No, I am not consulting with the ghosts of Christmas past or Christmas present they do not understand binary computing, or even virtualization.

Sure, virtualization has changed server side computing, sure virtualization is changing desktop computing, heck, virtualization is changing cell phone computing, yes, virtual machines on cell phones, of course. But in the rush to improve computing utilization, have we missed something? We have, and we have missed it in a big, big way. Clouding or utility computing, depending on how your organization defines it, is wonderful, the SETI project proved that distributed computing, or floating processing is viable, and functional, so of course the idea of floating resources between datacenters is an easy goal to promote, no? And virtualization makes it, at least from a tactical perspective, possible, and from a strategic perspective, logical. But what did we miss? Still guessing?

No it is not application instancing per se, although I have long said that application instancing is going to be a big deal, a least a few times in this blog as well as elsewhere. No, the next big deal is operating system reduction. Gartner do you have your ears open? What? What the heck are you smoking? I am sure some are saying…You stuff a bunch of mistletoe in a pipe and smoke it? Virtualization does not reduce operating system instances, but application instancing should or does if done right. But that is not what I am saying, I am saying, operating system reduction according to type. Yes, type. How many organizations are struggling with multiple operating systems while trying to craft a cloud or utility infrastructure? Why do multiple operating systems exist? Well that is a complex question and a straight forward question? The answers are complex and straightforward as well.

Let us take the simple question first, why do multiple operating systems exist? In simple prose, if you wanted a specific feature or variant of performance, you needed the right operating system. Well, those days are gone, at least in distributing computing, there is no reason for any organization to have more than two operating systems, if that, one for servers and one for desktops, done deal. Why? Simple, it is not worth the resources and expense to maintain a multiple operating system scope and scale engineering and operational support models for each operating system. You can do everything on every operating system within common reason today, the performance differences are disappearing for must applications. Virtualization has encouraged this, by further abstraction of the hardware. After reducing the total number of servers, and reducing the total number of operating systems, increasing the utilization curve, what is a big cost issue? Personnel, yes, even in these hard economic times…I heard the word depression echoing in the vast, empty, cube spaces of corporate America, no? It does not make sense to have a UNIX team, a Linux team, and a Windows team, now does it? Never mind that many large enterprise firms have several different variants of UNIX or Linux and even Windows running of very long periods of time? Why does it take so long to certify solutions? Why do cross operating system solutions fail to work well? You avoid these issues, when you eliminate operating system types. Imagine if you had one and only one engineering team focused the one operating system? That would save personnel expense, because yes fewer persons, but more important, less time to get to solutions to your clients, since only one platform is focused on, be it Solaris, Red Hat, or Windows 2008, to be specific via example.

Let us now tackle the complex question, how many organizations are struggling with cloud or utility computing when multiple operating systems are available or viable in the enterprise of the organization? Cross operating system solutions never do well consistently across all operating systems, in fact, they often lose focus and support all operating systems according to lowest common denominator. Why? Oh, it would be easy to say political turf, right? And that is actually true to some degree. There is absolutely no synergistic effect when you are trying to be all things to all platforms. At some point someone at the top has to make a hard decision and stick to it. This is both true for the developer of a solution and the users of said solution. But the greater cause is avoidance of pain, it is much easier to do what has always been done, because UNIX or Linux or Windows has always been there, it is easier not to force change.

Even virtualization is nothing new, just a slight improvement on the original problem, supportability and utilization. Virtualization has improved utilization, but made supportability worse, or should I say more complex? Absolutely everything in the information technology scope is complex, more complex than it needs to be. Why? Because marketing loves new features, and suckers, I mean upper management types find it every easy to fall in love with a specific concept or feature that makes a quick bonus, then standing up for a long term, far more superior objective, which should be more profitable years down the road, but then again, no high end of year bonus? Tell me I am not right? Has the number of vendors demanding to present new toys to your engineering group increased or decreased over the last year? I for one, have never seen so much junk, yes, junk, solutions flood into by electronic mail, snail mail, or even the door to my office, because there is a hard blitz on to chase the latest new toy. When did early adoption become strategic planning? And of course absolutely every new toy must be implemented in less than 12 months, or else.

We need to de-complex the computing infrastructure in order to create a new cloud or utility computing model that will give us long term, or more than 5 years is what I call long term, and significant strategic benefit. The first step to that goal, is don’t implement every operating system in the cloud. And if not implemented in the cloud, let those orphaned operating systems decline, and disappear. Until this is done, cloud or utility computing will have a flawed foundation, a weak point, that will drive up cost and complexity, which is not acceptable. Cloud or utility computing must be easier than ever before, must be leaner than ever before, it can not survive the silo effect of conflicting operating system goals, objectives and needs. Cloud or utility computing should be monolithic or it will just fail to live up to even modest expectations. So, until the number of operating systems up for consideration in your cloud or utility model is reduced, I say, bah humbug. Oh, and the ghost of Christmas future, it, agrees with me…Ha! So stick that in your pipe and smoke it.

Personal note…To all that read this blog, including those that have agreed, disagreed, and stated such in comment or reply to this blog; I wish you all happy holidays for 2008 and best wishes for 2009. I hope that you have enjoyed this blog, and will continue to enjoy it next year.

2 comments December 17th, 2008

Virtual Instance Performance Revisited

Virtualization, Fine, Well Sort Of? – Chapter 08

This is a revisited article, not because of a correction or change of view, but advance a topic that I have always indented to revisit, but never seemed to have time to do so, until now. A loyal reader of this blog reminded me of this fact recently, so I am honor bound to resolve the gap or lack of continued discussion on this topic. Oh, I am speaking of virtual instance performance of course, as the title notes. Unfortunately, virtual instance performance is a complex topic that gets into the swampy weeds full of tangles and hidden snags, faster than water and dirt make mud.

In part I of this topic, I discussed the context of performance, to which, Peer-to-Peer or Inter Virtual Instance Performance, which is what the host infrastructure reports. I will not rehash this topic here, but it is important to note that only the host infrastructure can accurately report virtual instance performance. Also in part I of this topic I referenced, Host to Instance Performance, Host to Host Performance, and Host to Cluster Performance. Respectively I will summarize each concept, but for more detail, refer to part I of this topic. In brief, Host to Instance Performance, is overhead of the host or host impact to performance, what does your hardware and hypervisor cost in reference to performance. Whereas, Host to Host Performance is which host executes which virtual instances the best, all things beyond individual instance deltas being equal. Moreover, Host to Cluster Performance is one step short of cloud or grid computing modeling, focusing on which host in a given cluster is the most efficient given a known set of virtual instances. This is important, when you consider data center globalization, to which hosts should be consistent, and so should clusters of hosts, across different datacenters, for example.

Now if you are tracking all these This-Versus-That models above, then you will realize that one model is missing, which it is? Give up? Cluster to Cluster Performance! There is a good reason for this; I neglected it in part I, my bad. As life cycle and management tools have approved over the last year or so, this is a viable and significant performance model, especially when you have heterogeneous hyper-visor based environments. Consider VMware versus Hyper-V, or Xen versus VMware, or Xen versus Hyper-V? Obviously if you offer a class of service aspect to your virtualization, you need to be able to compare different virtualization infrastructures in real time, with little or no explicit normalization. I for one, hate normalization, it is often abused and biased to a specific or narrow criteria set, so normalization devalues the analysis and results. But I digress, Cluster to Cluster Performance is beyond the scope of this specific article, but will be discussed in the future, did someone say Virtual Instance Performance Part III?

But the title is Virtual Instance Performance Revisited, and so the key to all performance evaluation starts and ends with the virtual instance, this is the corner stone of virtualization, be it application instance, virtualization container, or operating system isolation based. The vast majority of tools available for virtualization performance evaluation focus on the virtual instance of course, since the goal is, to always have the fastest instances possible, given the constraints of the associated infrastructure. The last comment begs the question, what constraints? Well, these are discussed extensively by virtualization gurus over and over, including processor context switching or processor cycle loading, memory IO, disk IO and network IO. It is quite common for various hardware vendors to focus only on one or two of these constraints and publish misleading or flat-out inaccurate statistics declaring they have the best or fastest virtual instances in the known universe, only on their respective hardware of course. Bah Humbug! They even normalize their results in comparison to their competitors to prove the point that they have the best hardware. Bah Humbug, Again!

Unless you evaluate virtual instances only under severe load for all four (4) constraints, and inclusive of these constraints, you are not doing your clients or yourself right. A classic garbage-in gospel-out (GIGO) scenario if there ever was one. Virtualization abuses hardware, it is and always does this, this is by design, and after all, virtualization is attempting to fully utilize resources that are often unused or wasted, no? So the single most import issue with virtualization instance performance evaluation is the selection of the tools, not I stated, tools, to do the evaluation, Ah Ha! Bet you did not see, or read, that one coming now did you? VMware VMark, vCompute, IOMeter, etc., all have their weak points, you must understand these limits or issues before you design your evaluation criteria and methodology. Consider this, if your specific testing for virtual instance performance testing is only looking at processor loading and memory IO, are your clients not going to be unhappy when their network IO and disk IO results are horrible? Did you analyze your environment right? Did you evaluate your proposed environment right? If the virtual instance evaluation for performance is skewed, then your entire environment performance evaluation for Host or even Cluster scope performance will be horrible.

Now it is time to get into the weeds, and get mud in between the toes. Now that we know that we must test all constraints explicitly and inclusively, and we must test at the virtual instance scope before all else, what do we do? The virtualization gurus will argue over this, but below is what works for me.

  1. Establish a control. Establish a performance history baseline. If you are testing virtual instances on a new hyper-visor, or new hardware vendor, do the exact same test on an environment you already understand. If you have HP, and test on Dell, don’t normalize your results, just make sure you understand that HP and Dell are different, and make sound inferences based on the raw results. If you can test on 3 or more hardware vendors at the same time, or have historical data using the same tools and methods, you don’t need to normalize the data. Normalization is for management and others that do not know how to analyze resultant data.
  2. Processor and memory differences, including changes in caching speed and size of buffers, are often a shifted scale comparison, so normalization is not needed. This is also true of power consumption curves. This is rational and logical, since network and disk sub-systems should remain consistent for a longer period, so by definition the number of factors to be compared can be reduced if the sub systems remain consistent, including, of all things, the PCI bus architecture per host. Production performance data always trumps lab data. So if you have HP and Dell in production, and are evaluating IBM, use the production data as the baseline or control, then test HP and Dell with the newer tools or methods, or processors and memory, etc., then and only then, test IBM. Bingo! No normalization is required. I can just hear the slick stylized marketing types for all the various vendors crying over their iced-mocha-lattes, when they find out I always reject normalization based evaluations by default.
  3. Always run the same test, in the same environment, at the same time, with the same characteristics. This is just basic common sense. However, don’t be surprised when you see something that does not make sense. Iterations are key to the entire evaluation effort. Remember that basic statistical analysis requires that a sample size of 30 or more is needed to get to any standardization and variance deviations accuracy. Every time a change is done, changing the experiment is done, and performance evaluation is an experiment. Think scientific method all the time when doing any performance evaluation, be it in the lab or otherwise.
  4. Make sure you understand where and when you can introduce error into the results. The only way to do this is through peer review, getting more eyes on the proposed test plan, is the significant objective. Everyone sees the same process with or from a different perspective, whereas tunnel vision is evaluative death. Sometimes eating crow at the beginning is better than getting heart-burn while coughing up feathers at the end of an evaluation effort.
  5. Control expectations. Data often goes around the world faster then the executive summary. Expect that someone, somewhere, will take the evaluation tools and methods, as well as the results out of context. Results will be challenged, be prepared for it. Don’t defend results, only explain how results are generated and analyzed. Vendors hate this, and often forget this point, when they sponsor or quote so called independent analysis, focusing the resultant explanations as the authoritative final qualitative statement, when the raw data objectively discounts or obviously points to other conclusions. Normalization often hides the true results.
  6. The developer of the given virtualization environment is the start of the process not the end. Do not rely on the developer tool set, nor what a given vendor demands as the only acceptable tool for analysis. Of course the vendor has tuned the given tool or methodology to illustrate the strengths of the platform in question. Would it not be a wonderful world if HP performance tools worked on Dell and IBM, and Dell tools for same, worked on IBM and HP, etc., etc. Would make for some interesting evaluations no? Or Fabric tools worked on FCoE infrastructure, and iSCSI tools worked on FC infrastructure? Sounds insane? Not so. Generic tools sets exist, independent tools exist, use them. Even if every vendor in the world has used VMmark, VMmark means nothing to Hyper-V.
  7. Repeat, repeat and repeat, change only one thing at a time, for example, only change loading of one constraint at a time, be it processor loading versus memory IO, versus disk IO or versus network IO. Use the same dataset or streamed sequence for each test. Never change the dataset or streamed sequence between iterative testing for a given factor. Complete an entire set of tests before mucking with the variables beyond the planned test set. Could be considered a repeat of the point above, about running the evaluation in a consistent manner, but it is so important, it if it is a repeat, so be it.

Well, at this point, I am sure someone is yelling…But he has not told us anything useful yet? What the Heck?! Not true. It is true, I have not spelled out an explicit methodology for evaluation, as a do this, do that, then do this scenario. To do that, cough, would be to create a bias that should, no, must be avoided. But to be fair, I will summarize things a bit, and recommend a best practices approach.

  • Analyze the virtualization environment, focus on the virtual instances first, look at processor loading, memory, network, and disk IO loading, create and execute tests that stress all constraints as applicable to your expected needs, and well beyond your expected needs. If the majority of your virtual instances are encoding unique video data creating results, expect lots of disk IO, if your virtual instances are web servers, expect lots of network IO, etc. Be smart in your evaluation design for performance.
  • Remember, virtual instances are the corner stone of all evaluation, Hosts and Clusters have their own performance characteristics, but they are impacted or resultant based on the virtual instances. Dynamic resource sharing, high-availability, etc., are wonderful features, but mean nothing if individual and grouped virtual instancing performance is not understood. The goal is to have most of the virtual instances perform well, most of the time, nothing more. The number of instances, the number of hosts, the number of clusters, even the number of virtualized datacenters, if it comes to that scale or scope of evaluation, will be obvious and straight-forward, if the methodology and tools used are sound according to the virtual instance modeling.
  • Performance evaluation is a living breathing animal, and should be viewed as dynamic and experienced based, no pun intended. Nothing in virtualization is static, so allow and expect the methods and tools to be flexible and adaptive to the effort at hand. This is not to say that change is good for the sake of change. Only change tools and methods when it makes sense to do so. Never change technique in the middle of an evaluation effort. To do so is statistical resultant evaluation suicide.

Add comment December 10th, 2008


December 2008
« Nov   Jan »

Posts by Month

Posts by Category