HansDeLeenheer

PowerCLI for Dummies at VMworld

2013-04-24T12:19:00.002+02:00

Yes, I did it. I submitted a session for VMworld. And with that you also all know now what my new baby project is: PowerCLIforDummies.com. It's something that grew from a VMware Roundtable podcast every Wednesday and as usual when idea's grow, someone needs to take action on it. Because the idea was mine I took that action myself instead of leaving it out in the open.

So 2 things here: go to the PowerCLIforDummies website and follow or contribute my journey to become a PowerCLI Master - YES I USED STAR WARS REFERENCES!

Secondly: I submitted the website and the project for a VMworld breakout session. I will go through all the stages I am going through myself. The sesseion will go through the stages but will also just try to motivate people into taking on new technological challenges and how to get there. The Jedi Ranks I used really do make sense here by the way!

So, go and vote, you know I'll end up doing something hilarious up there :D

Title		PowerCLI - from Youngling to Master

Abstract		Learning a new technology is like learning a new language. It is especially so if you want to learn a scripting language. At the VMware Community Podcast of March 27th, where the topic was PowerCLI, a question got raised where you should start when you are a rookie. The answer was "many places". As a result of that PowerCLIforDummies.com got created that would be just that landing place to start. Making it interesting the website evolves through the stages of learning a new language. To keep it geek enough the learning process has been aligned with the Jedi ranks. By the end of this session you will know how to grow from Youngling to Apprentice, Knight and in the end become a real PowerCLI Master.

VMware's Software Defined Storage - and some rant

2013-04-11T10:59:00.003+02:00

There are 2 reasons for this blogpost. Firstly I was at VMUG Italy (#VMUGit) last week and the closing slot was Cormac Hogan (@VMwareStorage). Although he got the last slot, he nailed it! I will first give some hard facts about the presentation; what IS VMware's story on SDS today? Secondly I still consider myself an indy storage blogger so I do have some rant on the entire SDS BuzzWord. If you don't want to hear the rant, just stick to the first part ;)

VMware's Software Defined Storage:
At first Cormac starts with a decent disclaimer; everything we talk about here is projects and vision VMWare is working on and is in no circumstances a final stage or GA-ready product. In fact, there are no date's given, just vision and sharing at what stage in that vision we are today. Although some nay-sayers will have their own rant on sharing what without when, I do like the openess how Cormac presents it.

"VMware's Vision: Provide storage services & SLA automation for all applications across all types of storage". Off course this is a Product Marketing line but when read with human eyes I read: in the end it's all about the application and making sure the data meets a certain level of service based on automated profiling, no matter what the underlying hardware would be.

This does not mean the underlying hardware (hence the hardware vendor) is becoming obsolete. Quit the contrary; VMware (or other platforms) will create their "industry standards" against what the Storage Vendors can be compliant. Think about what VAAI & VASA for VMware or ODX (Offloaded Data Transfer) for Microsoft have already done. It's nothing more than offloading storage requests to the underlying hardware, but nevertheless asked for and automated by the platform on top.

The 3 (current) VMware Storage Projects:
vCloud Distributed Storage
Subtitle: Enabling per-VM data services on vSphere. Another name that circles from time to time is vSAN. This will use the local storage of vSphere hosts, SSD's + HDD's, and create a "Reliable Array of Independent Nodes", codename RAIN. Simply put a datastore will be striped over multiple hosts and then it also will have replica's of that datastore. The distributed array will provide it's own reliability and availability services on per-VM based settings. Today the number of nodes would be up to 32.

My 2 cents: Although it might be expected that Virsto will still be sold temporarily as it exists today, it will suprise no-one that a lot of it's technology will be part of vCloud Distributed Storage (I liked vSAN name better though). Another remark I have is if this will all be VSA based or if it would become part of the ESXi kernel. Although I would love it to be the latter I do question if this would not make the kernel too big again. Some people in the industry think this might kill SMB storage as we know today. I don't have an opinion yet on this. What I do know is that the 32 nodes scalability is really nice already. There are a lot of other vendors out there that struggle with big scale out namespaces, let alone trying to scale beyond that number in block based independent nodes. Note: this is not to be confused with vSphere VSA that today has a max of 3 nodes and is - in my opinion - a toddler product in the SMB storage.

vFlash
Subtitle: Enabling new tiers of storage on vSphere. VMware hasn't done a lot for Flash up to today. We had "Swap to SSD" in vSphere 5.0 and SMART (Self Monitoring Analysis and Reporting Technology) in vSphere 5.1 but nothing that really changes the way you write IO because of the fact that it is flash. vFlash will do just that. It will integrate SSD's, or any other local flash type for that matter, in to the vSphere storage stack. It will provide write-through and read cache software which would be transparent to the VM. There would be even an opening for 3rd party flash services.

My 2 cents: I like the fact that there will be an opening for 3rd party integration. Although I don't exactly get where this will be going from VMware's point of view yet but another example in the industry that I liked a lot was FusionIO's ioMemory SDK that enables applications to talk directly to flash storage. What I do question here is how much this would "hurt" EMC XtremSW Cache and others that have done exactly the same thing but the other way around (using flash as a band-aid for "too slow" legacy storage - pun intended :D ). Same question here is wether or not this will be embedded in the ESXi kernel. I guess this makes more sense if it does.

Virtual Volumes - vVOLs
Subtitle: Enabling per-VM data services on the array. Basically vVOLS shifts external storage granularity from a volume base to a VM based methodology. Today a storage array has disks that are combined in a storage pool (mostly some type of RAID) and then carved up in LUN's that are presented to the Hypervisor. In that LUN we put a lot of VM's with it's VMDK's. Sounds familiar, right? One of the features vVOLS will introduce is the Protocol Endpoint (PE). The PE will be similar to that storage pool in which the array will put vVOL's instead of LUN's. This way we can create reliability and availability services like replication and snapshotting into the vVOL instead of into the entire LUN/Datastore. Important thing to remember is that the PE itself will not service the IO, it will only point the Hypervisor to where the IO for that particular vVOL needs to be directed. This will enable some scalability we don't have today.
LINK: Virtual Volumes Technical Preview - Cormac Hogan, dec 18 - 2012

My 2 cents: this is the most exciting part for me as it talks about how the hypervisor platform engages with the underlying storage platforms from other vendors. This is where the SDS nay-sayers are proven wrong! It's not VMware that is selling SDS, it's VMware that is building the framework to offload any VM data operation to the storage. Guess what you'll need for this ... A STORAGE VENDOR! Early adopters will be on the winning hand here. I am thinking for example Tintri that already works on a per VM granularity but from within the array itself. They will be one of the first to be able to adopt VMware's vision and technology. Anyone else dismissing this will be "slacking".

---------------------------------- THE RANT ----------------------------------

Everyone has SDS?
important disclaimer: this is what I think today and I might change my mind on the topic as the industry and the buzzword evolves (covered my *ss nicely right?).

At first something about BuzzWord bingo. I get marketing and I get it that we need new stories to get new stuff sold. But just blurring out buzzwords because of the buzzword doesn't make sense. Making your own definition of a new buzzword is not a good idea either, as this only enhances the confusion for the actual customer of whom I sincerely hope we all care about in the end. Think about how long it took to be a little bit on the same track with the whole "cloud". And it still is not quit understandable for everyone. Another famous example in my daily work is Big Data; having/handling large quantities of data does not mean you are talking about "Big Data". It simply means you have a large quantity of data.

So on topic: a new buzzword does not mean you can rewrap yesterday's, or even today's technology in tomorrow's definition. Software Defined Storage as explained above simply does not exist today. Not at VMware, not anywhere else. The fact that you have a storage product that is sold as software and can be installed how you like it does not mean you have Software Defined Storage. It simply means you have Software that you can install how you like it and serves the purpose of Storage. Otherwise Windows Server 2003 with an SMB1.0 share is also SDS, and that is even a very recent example :D . I know that some analyst firms (IDC for example) are helping others to call - any storage that runs on x86 platforms - Software Defined but I have to disagree with that. x86 celebrates it's 35th birthday this year ... The difference for me is that Software Defined as it is presented by VMware today is Technology and Innovation driven and the other definitions are Sales and Marketing driven. A new technology does not really have to de dictated by a single party involved (VMware) but at least let's focus on the right drivers for it.

I know I might step on a few vendors' toes here that are preparing or launching there marketing machines for SDS but hey, everyone still has an opinion right? And please provide me with feedback in the comments. I love to be proven wrong on a daily base! It means I am learning and growing.

Learning new languages 2.0

2013-03-17T10:00:00.001+01:00

I was never good at foreign languages in school. I failed my French classes each and every year and my English was as good as I could learn it through what I got from Hollywood. This means forget about s/z differences (pun intended) and please don't get me started on Future Simple or Passé Composé. 15 years later I do mange to communicate in both those languages to the level that people actually understand what I want to say and vise-versa. My Hollywood English off course is still better than my Bistro French but it took me 15 years to get here just through exercise and not by learning as such.

Now, those 15 years later, I am in a professional position that more languages in the tool-belt are becoming extremely important. Being an EMEA (Europe, Middle-East, Africa) Evangelist for a software company means I get to deal with a dozen languages I don't speak today. German, Italian, Spanish, Greek, Russian, ... I think I covered the most important ones here. I didn't even touch the Nordics here (Danish, Norish, Swedish, Finnish, ...) but I guess you got my point.

In an effort to get closer to our customers, partners and even colleagues I decided to do something about that. It's going to happen and it can't take me years and years. At least 2, maybe 3 extra languages should be a 2 year goal. Manageable, right? Now we have 2 issues here: I suck at learning and I have no fixed week-schedule due to my traveling. So even if I would be able to be a good student, I wouldn't make it to class half of the times necessary.

The power of the masses:

Recently I downloaded the App "The Human Face Of Big Data", that presents a lot of stories how Big Data exists in the real world. One of them was a TEDtalk by Louis von Ahn, the inventor of ReCaptcha. Summarizing the talk in 2 minutes is pretty difficult but I'll give it a shot: there are millions of people a day using 5-10 seconds of their time to authenticate themselves as a human being on thousands of websites. This is at least 10.000.000 seconds together (=350 days of 8 working hours)! With those seconds today we are all digitizing old scanned books word by word as the human brains is stronger than any OCR program out there (Optical character recognition). Watch the video below for a more detailed explanation. This was where Louis von Ahn found the "power of the masses".

So what else can we do with millions of people? Let's change the paradigm: what can't we do due to lack of those resources today? One of them is translating the internet. Have a look at Wikipedia; only a really small percentage of the pages in English are available in Spanish. So how do we translate all the rest? Throwing it through Google Translate won't help that much right? And in fact it's the same issue as with OCR: computer have no context! So we will need hundreds of very expensive translators and a lot of time. Unless ...

DuoLingo:

DuoLingo is the next project of Louis von Ahn. It's an App that you can download for free to learn new languages. It's very user friendly, a little game-like with points and levels and the best part is: you learn a new language at your own pace for free! Is it completely free? Not exactly, you are paying with your time. And how can you monetize time? After a few levels, when you master the basics of the language, the app will start giving you sentences of foreign websites to translate in the language you a learning. If 20 people translated it the same way, the translation will be more accurate than any computer program could have ever done it. Multiply this single action for hundreds of thousands of users and here we have "the power of the masses" massively translating the internet.

The app also is a science project as such in learning new languages: it captures what works best to learn new words and grammar over all these users and changes the apps behavior to your personal profile. It will alter the exercises for you differently than for someone else because your profile seems to learn faster through another type of exercises (ie more visuals versus more auditive exercises).

The bet on technology:
Two weeks ago I visited our head quarters in Russia. One night I was telling the aforementioned story and my strong faith in this technology. My Russian colleagues don't believe in it and swear by instructor-led courses. Today Russian is not open in DuoLingo yet but I did take a strong bet on the technology: from the day that English/Russian is available on DuoLingo I have 6 months to communicate fluently in Russian with a total stranger.

Want to know more about an even bigger bet? Wait for my next blogpost! It will reveal the first language I'll take on as of today and the big bet behind it with another Language Learning Company!

Your storage in the Cloud

2013-01-27T15:54:00.003+01:00

Public Cloud, Private Cloud, Hybrid Cloud, bleh bleh bleh, ... get over it, it is all called CLOUD and it will remain that way for a while. What I haven't talked about yet is CLOUD STORAGE. What is it and how can you and I use it? I'll try to get to a basic overview of what it means and how you could get started using it. As Cloud Storage will become much more important towards the future I see some blogposts coming up later this year with specific implementation possibilities.

WHY?
Why do we care about storage that is not in our own control? $$$ might be a first and very good reason. Failover is a second one and sharing it is another one. Let's start with that first reason as I guess it has been the most important one. For a long time now we got away from the local storage and have put all our data in Block Storage devices (some call these SANs but that's not a proper name for it) and File Storage devices (NAS). The biggest benefits here were high availability (shared storage) and more performance (more spindles = more IO). But with the increase of requests for volume, these devices tend to be a lot more expensive. So the price of just storing data has tremendously increased although for a lot of that data you don't need such expensive storage. In come the Cloud Storage Providers.

How can they make storage so cheap? Well, you know it or not but those storage devices you buy today are waaaaay too expensive for what they are. A lot of the costs of your array goes to margins. Margins for the disk manufacturer, distribution channel or just ridiculous maintenance licenses. And the ones with the smallest margin in the whole story here are the integrators taking the time to install it decently for you. So how do these cloud providers make it so cheap? They build it off course. And they take down all unnecessary costs. And it's volume. Add all up and you can make your own storage pretty cheap.

WHO?
The most common known cloud storage today is Consumer Class: Dropbox / SkyDrive / Google Drive / BitCasa / ... that give you a fair amount of free storage. Note: if a service is free, you are the product! There are also a few fairly low priced alternatives for just that bit more than home usage such. I use ADrive for example that gives me more than the capacity I get from those Free providers for a fair price.

If you stretch the idea of these types of Cloud Storage one could go create it's own Ghetto Cloud Storage. I wouldn't go there myself but the idea is nice :-) Let's focus from here on to what is considered to be Enterprise Cloud Storage. Who should be in this list (for now)?

One of the things we heard over here in Europe is the concerns about the Patriot Act. Does the Government have the right or not to look into private data? There are not a lot of non-US providers and let me assure you: if the US administration wants your data, it's going to have it. A while ago blogger Aidan Finn wrote about it here. What we start to see now is that small local initiatives have taken that concern and made it a business case. Today I accidentally found a BeNeLux company (Green Storage Cloud) that does nothing else than store your data within the borders. I'll leave it up to the reader to pick their preference.

HOW TO USE IT?
One of the easiest ways to use Cloud Enterprise Storage is a Cloud GATEWAY. You have appliances like TwinStrata or an easy-to-use Software solution like CloudBerry. Normally these appliances do not have on-site storage. They are more or less comparable to proxy-servers for IO.

Logical example of Cloud Gateways process: TwinStrata CloudArray:

Amazon has also it's vey own gateway as a virtual appliance in vSphere (OVA format). Read the review here from Chris Evans (@TheStorageArchitect)

A second type of Cloud Storage is StorSimple, recently acquired by Microsoft. This appliance is the middle way between local storage (SAN/NAS/DAS) and the cloud. They use the principle of tiering and just use the cloud as the last tier. Let me show you how this happens in a picture:

Red: Flash / Green: SAS / Blue: offsite

So ... this appliance shows NETTO 100TB as shared storage on site. But in the appliance we only have 10TB of local storage NETTO. Those 10TB are deduped storage on 2TB of SAS/NL SAS disks and 0.4TB of flash disks. And the big blue part on the right? Yes, those tier 1 public cloud storage vendors. This was really a unique player in the market and the acquisition by Microsoft will definitely disrupt the market here.

The 3rd way 'out' is when your application can write directly to Cloud storage through their http API's like REST. Or you could even upload content to Amazon S3 through a simple web-form. If you write your own software there are also Amazon integration Developers toolkits from Force.com.

BUILD IT?
Next to building your own storage array with off-the-shelf components and open software you can also build your own Cloud Storage. VMware did an attempt in starting an Enterprise alternative for Dropbox which was called project "Octopus". At VMworld 2012 however Octopus had been moved to the "Horizon Suite". The Horizon Suite is where VMware has put some features together that are tied to End-User-Computing. Simon Seagrave drill the suite down for you: read here, I asked the Horizon team what this means for the product as such:

Is Octopus, now known as Horizon Data, only available as a feature of Horizon Workspace? - A: Yes
Does Horizon Workspace only work on mobile platforms (iOS/Android)? - A: No, there is also a desktop client for Windows and Mac.
Is Horizon only a object base platform (ex DropBox) or are there possibilities to share the data through file protocols (NFS/SMB)? - A: There is no intent today to use the Horizon Data platform as a file services. The opposite does exist "on the radar": a sort of Enterprise Gateway for existing NFS/SMB stores.

Bottomline: you really could build your own Enterprise Dropbox based on Horizon Data. I'll definitely will try this out myself later.

Others that play in the build-it-yourself area are for example EMC SyncPlicity & Egnyte. These are 2 solutions mentioned by Ray Luchessi in his latest blogpost: "Enterprise File Sync". I have no knowledge today on these products so I'll get in there later this year.

PITFALLS: The most classical pitfall of all times: SLA (Service Level Agreement). I'll give you a small exert from the Amazon Glacier SLA. Amazon Glacier is the CHEAPEST storage you will find today ($0,01 per Gb per month). The SLA however is open for discussion or at least leaves all responsibility open towards RTO (Recovery Time Objective). This is how it sounds: "Retrieving archives from Amazon Glacier requires the initialisation of a job. Jobs typically complete in 3 to 5 hours". Did you read the same as I did? What if it took 10 days? ... I did not say that these services are unreliable, just saying that you have to know what you buy! The information (data) is YOUR responsibility even if you signed and paid for a contract with someone else to store it.

PAY AS YOU USE:
A common factor amongst cloud storage providers is the way it gets invoiced. The price per Gigabyte is very low but that is only the cost of storage. The real cost lies in the handling and transferring the data. Uploading or downloading 1 file is a "request" and gets paid for. Very little but it gets paid for. And the second thing is the transfer. The amount of data transferred per month is also paid for. If you add these two things up it's actually pretty smart: either you'll have a lot small files or you would have a very small amount but very big files. In both cases the provider gets paid. Once the data sits offsite it is very cheap. This model is IDEAL for storage you hope you'll never need :-)

MY VIEW:
Keep a close eye on Cloud Storage! I know I will. There is a lot of data that I even personally have and don't want to lose. Why not put my data with 2 different Cloud providers? If I would do it for myself, why wouldn't a company? Seriously, think about this.

So as you might notice, this was not a deep dive. This was my knowledge and starting points as of today and I'll get through some/a lot of it this year. Maybe my 2012 was Flash Storage and 2013 will be Cloud Storage.

2013 - the year of FLASH in BeNeLux

2013-01-24T17:40:00.003+01:00

First post of this year and the look forward comes out of the look backward. You have seen me writing a lot about storage (flash) startups lately and I still have a few up my sleeves. But lately something odd has come to my attention. Everything has to do with the move of these (flash) startups to EMEA. Although you might say it is logical that some of these are coming over here, it appears that almost all of them use BeNeLux as a first base or at least second base (if UK is first base).

Some of the startups that I know making the move are: Nutanix, Pure Storage, Nimbus Data, SolidFire, Whiptail, ... I asked a few of them some simple questions on why and why this way:

Nutanix:
Nutanix is the reason I put the "flash" between brackets as they are a startup that do use flash as a differentiator but there is a lot more to Nutanix than the storage as such. Nutanix has build the "x86 mainframe" on top of VMware ESXi. Basically you really buy Compute / Networking / Storage in a box (wth 4 nodes) and you're done. Not enough? Add another member to the cluster. They didn't stop at 1 hypervisor though and already added KVM as a second platform. I heard rumours that Hyper-V is next (good move IMO).

My idea: Nutanix should get an as big as possible market share today. The biggest reason they are niche player now is that the server vendors have been lacking this vision. If HP would have bought FusinIO 2 years ago, this would already exist. If DELL would make something out of the RNA Networks acquisition + their cooperation wth Micron, this could already be in their hands and I am still wondering why Cisco UCS has not made this possible. So it will be up to Nutanix to take that x86 platform market now, because the others just aren't ready. If you want to talk about "Converged Infrastructure" - this is it, if you talk "x86 Mainframe" - this is it! Who's next?

Dheeraj Pandey - CEO Nutanix:
Finding early adopters and early majority in other parts of the world is better than locking horns with late majority and laggards in the US. We have an amazing Ops team that doesn't consider it insurmountable to support countries across the pond. BeNeLux and Nordics are some of the earliest adopters, world-wide. E.g., it was the most productive territory for NetScreen. VMware's two largest adopters outside US were Australia and Netherlands. BeNeLux folks are considered "neutral" VP's throughout Europe and in countries where the culture divide still persists. Leaders from there can be considered all-inclusive.

note: I love this quote! You see that this comes from an honest guy that doesn't feel the need to get all words being double checked by the legal or marketing department. But by being so honest you really see core values in his insights! Good thinking there mr. Pandey.

Pure Storage:
Pure Storage plays in the All Flash array market. I wrote a blogpost one month ago on the arrival of Pure in EMEA. And the basics of the product. In a nutshell Pure has made an array including the filesystem, from scratch, specifically for flash. They use NVRAM (today SLC) as a sort of cache before it gets offloaded in bulk (very small bulk but nevertheless) to MLC. On top of that they use a decent level of dedupe and compression algorithms. Pure Storage plays in the higher end of the current market where you would replace existing architectures for new. Pure has not really the intent to go after the top 1% HPC (High Performance Computing) but it would sometimes be possible according to the needs of a specific solution.

My idea: I can't hide my love for Pure Storage. They got the right vision, the right company mentality and the right product came out of that. I really hope that Pure get's those necessary early deals to get some traction. The only 'problem' I see is that XtremIO (now EMC) looks a lot like Pure Storage from an architecture & feature point so I hope they can win the battle in enough places once that monster of marketing and sales comes out of it's current alpha/beta state (expected 2013-H2).

Scott Dietzen - CEO Pure Storage:
We are building our our team and partner network in Benelux and the broader EMEA to attempt to duplicate our success over here. Northern Europe has in my own experience been a great proving ground for innovative new technology. And frankly it makes sense for us to move quickly while the competition is still trying to figure out how to copy what we have accomplished.

Nimbus Data:
Nimbus also plays in the All-Flash space with a high bandwidth, unified storage solution. Unified means it provides both block based protocols (iSCSI - FC - FCoE) and file based protocols (NFS / SMB). If you listen to Nimbus speaking, they really go after the top 1%. Performance first. This does not mean they don't think about the economics of the solution. A key point in the economics of Nimbus is that they make their own flash disks, reducing the margin overhead of the current flash disk manufacturers.

My Idea: I'll try to be as positive as possible. Nimbus Data has some challenges and most of them are tied to the fact that it still is a one-man-show around Tom Isakovich (CEO & Founder). I consider Tom as highly intelligent business man but I lack a few key points that they need to get Nimbus to that second level. Another point from lately is that Nimbus wants to go for IPO later. I don't know if they really have a product for IPO or if that is a move to getting acquired sooner. Actually maybe an acquisition would make sense to ad some senior company insights to the product. They are in this blogpost because I knew Nimbus hired an EMEA Manager before the summer of 2012 in BeNeLux. Recently I heard that this one guy already left the company and I have no news if this has changed already. This might be the reason why I didn't get an answer to my request for a quote.

SolidFire:
SolidFire plays more in the large datacenter cloud area so is not necessarily a daily competitor to the former players. It's a very scalable solution that expands up to hundreds of all-flash nodes with millions of IOPs. They don't target however towards HPC systems but more to very differentiated workloads on one system. Not 1 million IOPs to one Oracle but 1000's of IOPs towards 1000's of applications. The backend is a distributed iSCSI over 10GbE. Recently SolidFire hired Tim Pitcher (former VP at 3PAR) as VP of International, based in the UK.

My idea: I have to admit that I haven't seen the product as such. I do however support entirely the idea behind multiple workloads on scalable systems versus very high IO for single applications. The only thing I should look at is how they have countered the issues you see in some products when you scale iSCSI that much. Highly scalable products mostly are based upon namespaces or other protocols such as AoE (Ata over Ethernet) from Coraid. What I do know about SolidFire is that these guys are down to earth and really honest. This is what we do, this is what we don't do! Good mentality!

Dave Wright - CEO SolidFire: We have sales engineering resources in the region and will continue to grow the team in 2013. Our international expansion has been driven by strong demand for SolidFire's all-SSD storage system from cloud service providers outside the US, including 3 European service providers that we announced as part of our GA announcement (Calligo, CloudSigma, and Databarracks). Working out of our home base in London, we are currently serving customers across Europe, including the UK, the Nordics, Central and Southern Europe.

Whiptail:
Whiptail plays in the high bandwidth, high IOPs, low latency ... you get it, that top level performance players. High performance DB's, HPC, VDI (off course) but also low latency solutions like trading applications. The difference with whiptail is in the architecture: although it is an array handled by 2 controllers, these controllers only do IO redirection. The actual ACK happens in de storage node. How's that possible? Every storage node has a compute part so these are not dumb JBODs. A direct result here is the need for very high bandwidth, hence the 40GB IB backbone. Putting 1 and 1 together this is a linear scalable platform with a 0.2ms latency. Not enough? There is also a one-unit solutions with no IO redirection which results in a latency of no more than 0.12ms. The fronted is FC - iSCSI - IB - NFS. (note: where is SMB3?)

My Idea: here again you have a product that is well thought. It is slightly different from others. Again like mentioned for SolidFire, real scale-out means that you are mostly dependent on the scalability of the filesystem or the sprawl of connectivity. Something we have seen being handled by distributed filesystems with big namespaces instead of block storage. Here we see a best of both worlds. Whiptail is in it's C-round of funding so there might still come a 4th round. Nevertheless they managed to already have 25 people active in EMEA (mostly from the UK), some of them for 2 years now and recently they hired 3 BeNeLux people. I think chances are big you'll find a blogpost dedicated to the Whiptail Tech later this year.

QUOTE: expected shortly.

Nimble Storage:
Nimble Storage is also considered being a flash startup although they are not an all-flash solution. Nimble only plays in the mid-market specifically in the replacing current infrastructure place. The basics of the product is a Hybrid (SSD+HDD) array. CASL (Cash Accelerated Sequential Layer) is the filesystem they designed which should be a smarter model than what the current bigger players are doing by just adding flash as a higher tier or putting flash accelerators like FusionIO or LSI/Micron PCIe cards in front of their spinning rust (pun intended :-) .

My Idea: As mentioned before Nimble doesn't really play in the high performance market so has a much bigger install base to go after (NetApp/Compellent/VNX/...). Nimble could be in any infrastructure replacement deal. Because of the limited amount of flash (as an expensive component) combined with the compression, these products already are competitive today. I added Nimble to this post because they were already in EMEA in the UK, Nordics, DACH and BeNeLux. Like Whiptail they also have a few dozen (I am not allowed to disclose the real numbers) implementations in EMEA.

Philip Turner - VP Sales EMEA Nimble Storage: Nimble is investing heavily in EMEA and has established sales teams across Europe. Our Headcount is projected to grow in all geographies through 2013 following consistent, stellar growth throughout 2012. ... There will always be room for all-flash niche players at the upper end of the market to cater to the 1% of that market but using long proven technology (spinning disks) combined with backup and replication techniques (redirect on write) add up to value proposition that most people seem to have no trouble grasping.

note: it's a sales pitch quote here, that originally was 4x as long and says not much about the C-level vision. But it gives a clue what Nimble is about: they don't need the real flash fuzz as they are already in the current market replacing existing infrastructures.

WRAP UP:
Ok, we've talked about companies and technologies. But the intent of the blog was to point out that 2013 will be the year that all these startups are coming to EMEA. More specifically all of them with BeNeLux teams. These are my take-aways:

It's going to be a crowded market in BeNeLux. Although Northern EMEA has been known as an early adopters market, it's a small market! I expect all of these players together on multiple offerings. Even beyond the initial sweet spot of their products (which is not necessarily a bad thing).
Because of that crowded market, all flash vendors should be having a real EMEA strategy, not just a "we'll start here and grow from there". Those are doomed to fail!
Because of that crowded market, early adopters will get big discounts and huge amounts of support! If you are looking at replacing datacenter infrastructure or if you have specific requests that require flash, make sure these guys are on your list. You won't regret.
ProTip for the startups: don't forget about the big boys. Although you probably kick ass they will definitely kick back and they still got leverage in the customer base. Be specifically aware of EMC in Q4!
It is going to be difficult finding enough people with decent storage knowledge. As we see more storage players recruiting, there is going to be a point where the level of those recruits will drop. We simply do not have enough skilled storage people over here. Education will be key!
Channel is always the best way to sell solutions but be aware that channel (including distri) can't handle more than 1 or 2 players in each market. Looking forward to how this will get solved.

EDIT - January 26:
It has come to my attention that I missed a few other players that also are active in BeNeLux (active= recruited local sales & engineers). I will add their names here as they come in; Tintri, GreenBytes, Violin Memory, ... I'll try to find a way to provide contacts of all companies mentioned for Channel & Customers later. Note: I personally have no benefits from doing this, except maybe some extra respect from the startups.

Disclaimer: some/most of these vendors were at TechFieldDay last year where I was invited as a blogger. My travel and accommodation was "taken care of". This does not mean that I get paid to write about them.

2012 was Squirrels and Bacon

2012-12-20T13:14:00.000+01:00

2012 has been, without a doubt, the most exciting year since I walk around this globe. There have been other years that had higher or lower peaks but this one flipped everything upside down. Up to this year things have gradually grown on me like every "normal" person has. I changed jobs a few times, got up the ladder a few times in that process. Things just went fine for me.

Squirrels:

'Fine' is not really the best description here. I quit school at 18 and went to work the day after. I had 3 different jobs in the graphical world, all with the same employer but each time one step up. I followed a 4 years Networking Engineer Bachelor program in adult school (quit right before the end of it for personal reasons) and then had 4 different jobs in IT with VARs, each time 1 step up again. Starting to see a pattern here?

Something changed this year. As a Solutions Team Lead I had more responsibilities as before. Until that day someone else was responsible for my schedule and even for handling mistakes I would have made. Although my, sometimes very 'out-of-the-box', designs were good, I never got my files correct and on time. When you are in a (pre)sales supporting role this is very crucial to your job and the company. This is called procrastination :-)

At a certain point one of my managers, whom I truly see as a personal friend, gave me a tip that I should read some things on AD(H)D. I saw it first as a joke but he meant it and I got that hint. This has been a life changer since that day! I read a few blogposts and very shortly after that I noticed I had 99% of all symptoms to have AD(H)D. I bought books and finally I did go to a psychiator and psychologist to have it diagnosed.

I won't go very much into details here but I did hit a wall, and very hard, when this came reality. This explained 31 years of my life. But knowing now really helps. Small example: my wife knows now that if she asks me to do something and it doesn't happen right away it will never happen. We already knew that before but now she knows I did not forget on purpose, I was not lazy on purpose. So now she just asks it again without being annoyed and I get the questions more than once without thinking she's annoying :-) Other things we changed is that don't promise any works around the house anymore. We get things done now by others. Although that costs a bit more: it gets done! We live a lot happier now and that is a really good result.

I am still figuring out a lot and I might do some charity work in this area later. Talking at schools for teachers, maybe a TEDx talk. But I need some more research for that first. One thing that I already want to give to you all: DO NOT DISMISS AD(H)D AS A JOKE! It really is a bigger deal than you can imagine. I made that mistake myself.

Bacon:
A second big thing that changed was my daily activities. I really got sucked into Enterprise IT this year. This was not organic growth anymore. This was major leaps forward. I blog here for 4 years now, lived on forums and twitter (remember that 24/7 brain I have). I have learned a lot by just being there online so much, reading stuff and communicating with others. And then came the physical events. First I got invited by HP (dec 2011) to come over as a blogger to HP Discover in Vienna and in that same week the invite for DELL Storage Forum in London arrived. These two events were catalysers for everything that followed. The biggest catalyser after that was Storage TechFieldDay1. I still feel very humble being asked to join these grey-beards and brainiacs of storage.

But then came the ADD thing. I found out that I was failing in my role (as Team Lead). But after the fact that I knew now that just trying harder to be/look less lazy was not going to help ... I resigned. At that point the explanation was that the job did not fit my expectations anymore, now you know the other reason. I pursued the search for a role that fits my personality more rather than something I might be good at if I just try hard enough. I found that role as Evangelist at Veeam. These guys gave a crazy opportunity and it has been a roller coaster since I joined in august. I have visited half of Europe, have met a hundreds of interesting people AND I get to do something I truly enjoy: being passionate and talk to people about that passion. THIS is what I am really good at. I will never be a good salesman because I'll forget to call you back to close the deal. I'll never be a good architect because I will make smaller and bigger mistakes due to lack of focus. But I will be the most honest guy in the room as I lack the possibility of lying to you (I wouldn't even remember my lies). And I will be the most passionate guy you can put in a room and share thoughts with. Veeam gives me this chance every day now and I am truly grateful for that opportunity.

2013 predictions:
If you look back to all that happened in just one year I will not even try to do predictions for 2013. I can only tell you that I'll give myself for 100% in being passionate.

PureStorage arrives in EMEA

2012-12-08T20:00:00.001+01:00

If you follow my blog closely you'll notice that I like to follow up on some storage startups. A lot of them I did meet at a TechFieldDay event earlier this year, others I've met at tradeshows or other storage events. One of them that I truly love but did not blog about yet is PureStorage. If you meet these people you'll have to fall in love with them. The organization, the 'vibes' and not to forget: the product!

If I have to be very honest I have applied to PureStorage earlier this year when I was looking for a new challenge. They were not ready at that point to hire EMEA folks. A few weeks later some ($40M) Venture Capital fell in their mailbox but I was already on my way to Veeam, where I am now more happy than I could have imagined before joining!

But hey, times change and now they are here and it's time for me to give those credits they deserve. Let me start first with that company thing as such and we'll get to the technology later in the post. We visited The PureStorage HQ in Silicon Valley for the last session of Storage Tech Field Day 1. After 5 days on the wrong side of the pond I was extremely doped with a serious overdose of caffeine just to be able to follow until this last session. And it was worth it; we had Belgian style beers out of the coolest Symmetrix ever (video, 1:50) , we had Margerita's, baked bacon and some crazy sh#t, smart #ss developpers to present (ie. co-founder John Hayes). No BS presentations, no Gartner numbers or quadrants and certainly no marketing slides. Now 6 months later they hired the 3PAR ninja team! Don't get me wrong here. These are not HP guys trained to do 3PAR stuff, these were the real EMEA guys before they got acquired by HP. I have met these guys before and I must realy admit that there is no better fit for them nor for this company. Well done!

Flash well done:

We've seen a lot of different ways startups handle flash. The biggest difference between the startups and 'the others' is that these products are designed from the ground up for flash. It is not written for disks and have added flash to be faster than disks (ie VNX, EVA, Equallogic, Lefthand, ...). Will these products run faster than their predecessors; yes. Are you using that flash to it's optimal conditions? Not at all.

So how can you handle flash in a different way? You can use cheap MLC disks which gives you a decent $/volume but they wear out really fast when you hit them fast. You can use more expensive SLC disks that can run even faster and last longer and run some deduplication on it. Or ... you can be smart and use both for it's best conditions. PureStorage basically gets the writes in on an SLC disk and offloads them in batches to the MLC disks. If you look at their marketing it says 100% MLC but that is because the SLCs will be replaced in the near future with for example NVRAM. What happens is that 4k blocks come in, are grouped together untill you have 512k and is then offloaded to multiple MLCs. All of this goes extremely fast and is deduped inline.

One of the things that wears out flash disks is random writes. To do a write to a flash cell it has to be erased first before you can rewrite it with another value. PureStorage solved this by redirecting ALL writes to new blocks so that there are NO overwrites. So your system will be full very fast then? No, because PureStorage does garbage collection in the background but that is going to be a lot less than otherwise. One of the advantages here is that because the array already handles garbage collection they'll need less overprovisioned free blocks and can use more active cells in the SSD. Another advantage here is that you could for example do an un-delete on human error. If an admin deletes a snapshot, clone or even a LUN, you can simply undo that action as long as it has no been cleaned off course.

Smart architecture:

Everything is based on 'off-the-shelf' hardware. This does not mean off course that it's a whitebox everyone can build :-) It only means that there is no custom hardware like ASICs (ie 3PAR), customized SSDs (ie Nimbus Data) or high density disk shelves (ie Amplidata). The smart part of this type of architecture is that the technology is in the software and you can flexibily change that physical architecture. Remember my blog on the new bany 3PAR this week? The reason it took HP so long was the design of a new ASIC.

One of my pet peeves when I talk to storage vendors when they come up with new products is how you can move to a next generation. I do not like rip-and-replace technologies (Clarion/EVA/...) as they bring too much risks and thus you keep them alive way too long. One of my favorits here has always been Equallogic where you can just add a new model to the group and 'evacuate' an old model out of the group when it's time to retire. Here also Pure has been smart and made Stateless Controllers. In short this means that the controllers do not hold any data as hash-tables and such. All the actual data resides in the storage nodes. I once killed an entire EVA by clearing the controllers (I know, stupid no?). Whit this type of architecture you could easily fail and replace controllers or move on to another generation. Not all startups are this smart.

Features:

To be considered an Enterprise Flash Array just provisioning LUNs does not do the trick anymore. Thin provisioning, cloning, snapshots, replication, ... these are justs a few of the features. If you read back to what we said about writes always going to free blocks you already see at least two easy features to enable here: snapshots and cloning. As we are not overwriting data you can easily point a snapshot to those blocks that are still there. The only thing we'll need to do here is not cleaning them through the garbage collection.

Pricepoint:

So you are running MLC which makes it cheaper, you are doing inline dedupe which makes it smaller. How much does it cost? Well, the starting point for a single controller with a single disk enclosure is around $ 100k street price (add 20k for dual controller). You'll get a 5,5TB disk enclosure but it's only licensed for 2,7TB RAW. Basically PureStorage invests together with you (marketing, haha). They could easily have done the same with half the amounts of disks, like some other vendors do, but then you'll only get half of the performance. With an average reduction you'll get about 10TB of netto storage here so you'll end up at a pricepoint of $10/GB including maintenance AND ALL FEATURES ENABLED. That last point needs to be taken into consideration when looking at other solutions. Looking at performance this box should give you around 100k IOPs. This is mooooore than enough for most environments.

Wrap-Up:

This product for that price from this company ... where do I sign please? Seriously, if I had the money and I had a business case where this fits I would not hesitate. Let me know if you would and why

Links:

* As I was writing my post I recalled this technical deepdive from Nigel Poulton whas was also on that trip to Silicon Valley earlier this year.
* PureStorage website: great business case for PureStorage + VDI - read here

Please welcome baby 3PAR!

2012-12-04T15:59:00.000+01:00

Rebranding the 'future' story:

I've said it a few times in the last 2 years (actually since the acquisition of 3PAR) that HP should have replaced the EVA with 'newer' technology. FINALLY they made it there. So please welcome the new baby 3PAR: HP StoreServ 7000 series. You'll see a lot new rebranding stuff going on at HP storage. Let me give you a small rundown:

HP StoreOnce: deduplication appliances (HP labs)
HP StoreVirtual: fka Lefthand. Lefthand is now only used as the Operating system
HP StoreServ: 3PAR
HP StoreAll: IBRIX + Object (HP labs)
HP StoreEasy: Windows Storage Server
HP StoreOpen: HP LTFS
HP StoreEver: HP tape robots and libraries

If you read well through this list you'll see that a few products are not being rebranded. It's P2000 (MSA) and P6000 (EVA). This has a very good reason. They are already calling it yesterday's products and guess what; they're not making it in the future. My point last year was that HP should not have brought out the new P6000 because the only thing that was new there was the SAS backend instead of FC backend technology. There was no new things in the software what so ever. So yes, finally the EVA can go to sleep.

So what's new?

The new HP StoreServ comes in 2 models: the StoreServ 7200 - dual controller - and the StoreServ 7400 - dual or quad controller with full mesh connectivity. There have been some changes in the 3PAR technology that enables this to go to a much lower granular model than before. The problem getting the 3PAR to SMB market was the large granularity with its 4-disks drive magazines which also had to be vertically aligned throughout the array. I did a blogpost on that in december 2010. Now you can buy the smallest dual controller box with a minimum of 8 disks and upgrade per 2 additional disks. The only exception is on Nearline Disks where you'll need a 12 disks minimum because the best practise is to put RAID6 on it. So how's that for some granularity? And the official word this week is that the StoreServ 7200 comes in already at $25.000 list price.

Licensing:
A second issue HP 3PAR has been struggling with is the sprawl of licenses that go on top it. It still looks messy but at least they did a big effort in simplifying the story. So what happened? HP bundled licenses that are in a similar solution type into licensing suites. Here is an overview and some details on the new suites.

So let me help you to find a few important details in here:

There is a CAP on the disk based licenses and the cap is pretty low (for 7200 just 2 disk shelves already)
All the THIN licenses are now included in the Operating System Suite!
There is a license called Online Import License with a 180 days expiration. Actually this is a temporary license for the Peer Motion technology which is otherwise licensed in the Data Optimization Suite.
ALL licenses are still available in a single sale but the message here is that you get more profit of buying them in a suite.

So how about my grandpa's EVA now?

HP made the transition from EVA to 3PAR going really slick with that Online Import feature. Calvin Zito (@HPstorageGuy) made a nice video on it that he will publish anytime now so I'll make sure it get's in this blogpost.

Interesting Links:
Bart Heungens: New Features in 3PAR Operating System 3.1.2
Nate - TechOpsGuys: 3PAR: the next generation - good overall tech deepdive
Chris Evans - The Storage Architect - HP Storage bets on 3PAR
Robin Harris - StorageMojo - The Stovepipe vs The Grid - "...the only credible challenge to EMC in the last 15 years."

Are you ready for the Flash

2012-11-27T13:28:00.003+01:00

As promised at the last #VMUGbe (#17) I did deliver a session for the latest #VMUGbe (#18) november 9th. Although I did switch to 'the dark side' in between now, I do try to keep up with some independent storage talk. One of the things I have been involved with this year, partially due to the Storage Tech Field Day 1 trip, is flash storage. What is it, who are they, what are the differences.

Click on the image for the entire slide deck

Part 1:

Basics of storage placement, from DRAM to Cloud tape. The advantages of fast and close versus slow and far away. Cost versus SLA.

Part 2:

What is FLASH? SLC, MLC, TLC, eMLC and future products like ReRAM, NRAM, PCM, ...

Part 3:

Where is the Flash and who are the players.

In Server: DRAM: in memory databases (SQL/Hadoop) or L1 cache (Nexenta) / Server side flash: tiering versus caching ( FusionIO, Micron, EMC Lightning, ...) / new kids around the blocks: VNMe - SCSIe
All Flash arrays: commodity hardware ( Kaminario) / build your own eMLC / smart filesystem flash offloading (PureStorage) / Purpose build PCIe (Violin) / and many others
Tiered/Hybrid SAN solutions: legacy vendors (HP/DELL/EMC/...) versus startups. No RAID (NimbleStorage) / Dedupe/Compression & Offload (Tintri) / L1-L2-... offloading to disk (Nexenta) / Building blocks (Nutanix)

Part 4:

Virtualised flash solutions: VSAs could run Enterprise storage in local storage SSDs (vSphere Storage Appliance / Nexenta VSA for View).

Part 5:

What did VMware do?

Swap to SSD (5.0)
SMART - query your SSDs
vSphere distributed storage
from VAAI to VASA
and beyond to vVols

The home lab - general

2012-11-21T12:53:00.000+01:00

Finally I decided to create a decent home lab. Not only to run some tests but most of all to be able to show all of you in the conferences that I will attend reliable demo's. You know what happens when you want to demo on a over-used network ...

I'm not entirely sure if I will keep it running from home or if I will drop it somewhere in a hosting facility but for now it will stay very close to me :-)
Components:
2 x Server:

HP ProLiant ML350p Gen8 Hot Plug 8 SFF Configure-to-order Tower Server 652065-B21

Includes:

HP ML350p Gen8 Intel E5-2603
HP 1Gb Ethernet 4-port 331i Adapter
HP Smart Array P420i Controller
Eight (8) Hot-Plug (SFF) SAS/SATA Drive Bays
Seven (7) HDD blanks
Ten (10) USB ports
Eight (8) PCIe 3.0 slots (3x16, 1x8, 4x4), One (1) PCIe 2.0 slot (1x4)
4 x 16GB 2Rx4 PC3L-10600R-9 Kit

I also got 2 x Emulex 10GbE CNA's with iSCSI, TCP/IP and FCoE offloading. They also support SR-IOV so yes, very much looking forward to do some testing with these.

I added 1 WD Caviar Green 2TB per Host (WD20EARX) for the time that I would like to take the lab with me without the SAN.

2 x Switch:

HP V1410-16G Switch - Switch - 16 x 10/100/1000 - desktop. THese are very cheap switches. You will notice that this is a stron contrast with the CNA's ... I KNOW! But I wasn't ready to buy so expensive switches that would fully support them. Therefore I will try to use the CNA's as Peer2Peer direct attached ones and use them only as a 10GbE Storage backend. This should make it possible for me to run some different VSA's or even a Nexenta Cluster.

SAN:
Drobo B1200i; LINK
In this model of the Drobo we have 6x2TB disks but also 3x200GB SSDs. The SSDs are OCZ Talos MLC drives. I blogged about Drobo before with how the B1200i actually works. I am very happy that I will finally be able to test this myself in a real use case.

The Plan: blogposts will follow

Install ESXi 5.1 on first server
Build infra basics incl vCenter and distributed switches
Build Auto-Deploy for the Lab
Build nested lab for Server 2012 Hyper-V
Build fake production environment - the DEMO lab incl AD, Exch, SQL, ...
Throw all Veeam products at it!

TEDxBrussels - live blog

2012-11-12T09:09:00.001+01:00

Today I am joined by 2200 other TEDx followers. TEDx is an independent TED organization. Today we are together in one of the nicest venues in Belgium: BOZAR. Entering the room already gives you the energy you expect from a day like this. For the first time ever I'll try to live blog the event. Let's see how that turns out to work.

9:10 - Opening Choir: Brothers Kolacny and their 23 SCALA girls think we are "Fucking Special". Very sober version of Creep by Radiohead.

9:20 - Steve Wozniak: when kids go to school for the first time they are told that they can only learn what they are supposed to learn. Yet another great mind talking about the failure of our old-school educational system. The little boys/girls that can use technology as young as possible could be the innovato of tomorrow. The inhibitors of inovation are the big businesses and will assimilate the ones that would disrupt their big money machine. We need more of those little "builders"! Did you know the Apple I and Apple II were completely 'open'? 'Visical' was probably the first app that made Apple bigger because the Commodores and others could not expand for that workload. 20 years later the same happened by opening iTunes/iPod for Windows.

9:40 - Alan Greene: - Pediatrition - How can we transform the lives of children over the world? WHO (World Health aprganization) found that 29% of new born deaths are by now getting enough OXYGEN in their first minutes of life. So how do we get these children to get kore Oxygen in that room? Oxygen cans? too expensive. A second issue is that we have not enough IRON. Iron deficiency loses intelligence. Health and economic benefit would be oveqr 20% if we didn't have iron deficiency in the 3rd world. But this also is really hard to do and extremely expensive. the solution: at the moment a baby is born, a 3rd of their blood is still outside of their body. Leaving the "abiligan cord" on for an extra 90 seconds would give the children that extra oxygen and iron from their mothers blood! TicTalk campaign: changing the abilical cord clamping from "imidiate" to "optimal".

10:00 - Mitch Altman: #Wednesday. TV-b-gone. Big supporter of creating communities to create like HackerSpaces. what is a HackerSpace? it is a physical place where hackers can work/play together. Start by hacking yourself: learning and growing. Taking what is and sharing what you find. How about inventing a keychain that disables tv's? This was just a start. In the beginning a HackerSpace was only on conferences, after that it became more spread and now here are over 175 HackerSpaces around the world.

10:20 - Alexander Asseily: cultural explosions. Alexander talking about the power of drawings to explode culture. The Phonetinas gave us symbols - the phonetic alphabet - that helped us communicate. at that point in time those symbols were just abstractions of something real: "A" could have been the head of a bull wth horns upside down. The next big evolution was thousands of years later by Guthenberg: the distribution of word by printing multiple times. Radio and Electronic communication might have been the 3rd big explosion. Not only by distribution but more importantly about the access.

10:40 - Andrew Keen (@ajkeen): BANG BANG! Comes to talk about the future crime; The Digital Vertigo! What do you see when you look to the one in the chair next to you? Class, religion, food preferation, ... ? No, we see mysteries, secrets and questions. But more and more we just become data. We will see what people are about without knowing the real people behind it. Think about google glass, highlight, banjo, ... we are living the Truman Show. We need to learn to escape!

11:00 - First break: I am not blown away. Off course I truly respect the engineer within Steven Wozniak and the anarchism Mitch Altman breathes but the only one that touched me this first serie of sessions was Alan Greene. He showed us how a really small change in birth procedure could change the lives of so many people and in the end even benefit economy. This was an inspiring TED talk as I like them. Andrew Keen gave a really vibrant talk about closing up instead of opening everything to everyone. I respect his vision but as you know me by now, I don't really share his ideas.

11:40 - Zoe Laughlin (@ZoeLaughlin) : what are materials? Even the periodic table of elements is just a grouping of things that make materials. But are they? Even a copper spoon or a copper fork are different although they both share "copperness". is a 'crystal' the point of difference? Zoe shows us a single metal crystal of over 30cm that is used to make jet-engines. By the way, did you know Zoe has a moustache ... if you look closer wth a microscope :-) To see how vibrant she can be, look at this last part of her talk:

12:05 - Monte Stettin (link): presenting wth the iPad. #winning! Digital imaging can be far more powerful than simple text. But how to educate billions of children affordably? Monte showed us a lot of examples of the nice educational videos they are making but I do miss the part of how he is going to be disruptive and inovating. we already have some apps like his. Move it on Monte and go that one step further and change the world!

12:15 - Dale Stephens: really really really didn't like school! Decided not to go to 6th grade ... waaw! Left school to take charge of his own education. Volunteered in the community, worked at some startups in Silicon Valley. So he learned tons of things he couldn't learn in school and lots more. But besides the "hacking your education" life he had, it appeared that being gay had similar challenges. Who are the Velvet Maffia? The ones that are looking for benefits behind being gay? So Is there an advantage to being a misfit? And is that positive? If you embrace the fact of being vulnerable, you have a responsibility of being visible! Misfits are the X-Men of tomorrow.

12:30 - Eri Gentry (@erigentry): find your meaning in live and make somethng of it! We are only 10% human, the rest is microbes :-) but if you see that unseen, it's your doorway to discovery. Have a look at "DYIBIO", a group that started meeting each other weekly just talking about doing cheap research. The HackSpaces of Science. The community labs started to emerge. Eri could grow her company 'bio curious' from cheap to great in less than 2 years. Lots of crazy ideas could become big this way. It's not the PHDs that will change the world, it'll be the creatives that can grow themselves from the spare bedroom laboratories.

12:45 - Greg Gage (@gagegreg): I am a neuroscientist and spend 6 years in graduate school to learn how the brain works. That in such was already a shame that I had to go to grad school to get that information. So I decided to change that and did workshops in lower schools. but how to take $400.000 lab to a school? So they made "backyard brains", a small company that made a very small device that you can plug to your mobile phone and see what the brain really does. As a demo Greg knocks out a cockroach (puts it in ice water to sleep). After that he cuts one of its legs and put pins in that leg. These pins give us the opportunity to monitor the spikes that the neurons in that leg send to the central brain. And by giving vibrations back to the leg, it starts moving so we saw a cut off cockroach leg dancing to some beats. In the end he showed us the next part of the story: the "Cockroach Controller" a device so small that he can put on the back of a cockroach and control it's movements with a remote (aka smartphone).

13:10 - LUNCH: this series of talks was way more inspiring. You really feel that these are people that grew something from nothing, that try to change the world from their home office, think out of the box, disrupt the status quo in their own way.

Connie Hedegard - moved to afternoon sessions

14:10 - Bruno Zamborlin (PHD in Art and Computer Science - homepage): can technology help us understand things in the physical world? Everything you touch, makes sound, vibrates. What if we could extend its capabilities? Using the vibrations of a type of touch to do something woth these vibrations. Don't understand? Knocking on a teacup sounds diffent than tapping on it wth a spoon. Use the first sound to switch on/off the television and the second one to switch channels. And look at what happens if you hooks "objects" to a mini microphone that is connected to a cell phone:

14:20 - Aaron Rowe: every so often a shiney new technology comes along and big businesses want to grow on that. Urinal analysis strips are something like that. these strips can give for example pregnant women on a regular base if she has enough proteins etc. How about elderly in a home that get a box as-cheap-as-possible tests that show them the basics health of their liver. We made an app that can scan you tests and based on some of your metadata (age, gedner, ...) give you a better insight in your acurate health. How about upgrading those blood sugar meters and let them check more than suger only, hooked up to that app?

14:30 - Dr Stuart Firestein: it's very difficult to find a black cat in a dark room ... specifically if there is no cat. Scientific method versus farting around. "thoroughly conscious ignorance" (James Clerk Maxwell). Knowledge is big, ignorance could be bigger! Science is not a puzzle: a puzzle has been made and by design has an end. Science should always be wrong, it solves a problem by creating 10 more. So what we do with knowledge should give us stronger ignorance instead of the other way around.

14:50 - Tito Jankowski: co founder of Bio Curious and Scanadu, 2 companies we have already seen in the talks of Eri Gentry and Aaron Rowe. Didn't like this guy. Cheap sales talk from a guy that thinks he is Rock and Roll, wearing a Rolling Stones shirt and some John Lennon sunglasses on stage. this guy makes you think he bought this session (and maybe the 2 others too).

15:05 - Irene Van Peer: Product designer building 'stuff''. Can technology play a role between pigs and humans? "Pigs Chase" is a game between a human on an iPad and a pig with a huge touchscreen. You'll have to google this to see it. Very inspirational. For now it's only a concept video but the idea is awesome. Trying all different types of interchanging with these pigs really gave the scientists a lot of insights in pigs.

15:20 - Yasaman Sheri: as an industrial designer she thinks about the past, current and future of how specific objects are designed. There is a world beyond the physical what i called the digital. You can take a few pictures of yourself, load them through Autodesk and have it create a 3D model of yourself that could be 3D printed later by a webshop. The next generation of object designing will be about biological engineering.

Nice quote: "Whenever you feel sad, just remember the are billions of cells in you body and the only thing they care about is you"

15:30 - Peter Jansen: the fun of figuring things out. How if I could really invent the stuff in Star Trek? Take a "Tricorder". So I did. It can see atmospheric details, electromagnetic, temperatures, magnetism, ... And at a certain point in the designing process you feel the urge to share the knowledge you gained. So I made it an open source project. The morning after over 25.000 people had visited the website.

15:45 - Yvonne Cagle: former NASA astronaut. What made her wanting to go in space? And why doesn't everyone wants to go in space? This talk was more of a poetry session than a real inspirational TED talk to me. It seemed she still was somewhere in space. Would they have drugs for that at NASA? :-)

Just before the break we get an Australian student that won a prize with his PHD, Peter V Liddicoat. He made a super-alloy in aluminium.

16:05 - BREAK

MADE IN BELGIUM

17:00 - Jeroen Raes (@jeroenraes): people have 1,000,000,000,000,000 bacteries on your body. You are not human, you just a bunch af bacteria. But what if you start studying the gut microbiome wth metagenomics? They found 3 "entrotypes" like types of forest. These are groups that live together in a specific type of environment. This is the reason disturbance to your gut flora can give you diseases, diarhea, ... even obesitas can be caused by the disturbance of your gut flora. So don't mess wth babies flora. It can have a life-long consequenses. Some people can never recover by changing a diet because it could even disturb the body even more. Gut flora even influences behaviour and intellect.

17:15 - Ben Kestner (@KestnerTweet): breaking down the walls between schools and communities. If students can google the answer, we are asking the wrong questions. WAAW! amazing quote here. We are teaching to the test, not to the interest of the children. A school with no homework and no grades where you can learn what you want is more crazy than hoping on getting superpowers by eating an apple. Some of these schools are rising in the States. No tests, no grades and only mentors instead of teachers with a max of 15 pupils each. More than 98% of these kids get college acceptence. Think about that ...

17:30 - Rudi Pauwels: having a healthier life is what we wish for each year again. But how do we succeed in increasing life expectance? All that progress comes out of cost. And this will only grow. By 2040 we ar going to spend over 25% of our GDP to health care. Can we change that? Why spend billions of dollars on medicines that do not work for everyone? What if we could find a single moucule that tells us what disease you wpuld get? The future of health will be around focussing on acurate and more personal health care. Only then we will be ready to create "moleculair medicines".

17:45 - Alexander D'Hooghe: talking about "Universal Building". We started moving to the suburbs because they were green over 20 years ago. The result of that is that we attracted the supply chain industry to those suburbs. So we don't have green there either. How can we bound these thing together? People like nature, but we also like living in the community and a last thing we really like is progress. Combining at least two of these with each other in building projects. The killer could be having all 3. But what happens when you add all colors to each other? You get WHITE. Whatever you add to the white, it will get more towards one of the choices. So making white squares that can be filled in the moment and changed afterwards could be one of the nicest things you could design for everyone.

18:05 - Xavier Damman (@XDamman) CEO of Storify: regular people wth smartphones are the real reporters of today. Think about the guy that tweeted about helicopts in Afghanistan the night Osama Bin Laden was killed. WE ARE THE MEDIA! Speak up and let the world know.

At the very end I can close this extraordinary day wth Xavier Damman showing the room our own power:

18:30 - CLOSING DRINKS

Posted with Blogsy

Amplidata - Object storage well done

2012-10-29T13:19:00.000+01:00

A few weeks ago I visited Amplidata together with my friend Howard Marks (@DeepStorageNet). Amplidata is one of the very few Belgian companies that are a true Storage Vendor and it is the only one that has realized this with Venture Capital from 'the Valley' (ie Intel). So if people from the Valley are willing to invest money in a Belgian company it might be worth having a look. When Howard told me he'd visit Belgium for a project there I had no reason not to join :-)

What is Object Storage:
I personally categorize storage in 4 main areas: Block / File / Object / BigData. I categorize BigData differently just because it spans all of the others but however has it's own ways of dealing with the data (big data != a lot of data). The differences of the other three are based on the protocols used. Block: SCSI protocol(s), File: NFS/SMB(1/2/3/CiFS)/..., Object: API's (mostly over TCP-IP). In this case Amplidata uses REST API between the apps and the storage. Another aspect of Object storage is that you store the metadata of the file with the file and together this is becomes an "object". Example: a picture has multiple parameters like type of camera, gps location, timestamp, size, photographer, ... all this info can be stored with the file in metadata for intelligent querying afterwards without the need of a catalog/database.

Object storage is mostly used in larger environments and have a distributed type of architecture for protecting the data; holding multiple copies and longterm archiving for bigger files. One of the industry leaders in object storage is Caringo with it's CAStor product (also wth Belgian roots btw). CAStor has it's own distribution channel but is also available in OEM through A-branded vendors (ie DELL DX6000). Technically CAStor has a N-copies protection model. A file is stored on one disk and then copied to other locations: other disks, storage nodes or sites. Example: 4 copies of 1 object of which 2 on different nodes in site A and 2 on different nodes in site B

What's Amplidata's core technology:
The main difference between Amplidata technology and other object storage providers is that Amplidata objects itself are not replicated but the objects are cut in a lot of small pieces (chunks) with an advanced security math on top (Erasure codes) and then spread across multiple disks and nodes. Why this method? Because the security stays with the data itself, not by extra parity parts. Don't get it yet? Let me show you by example:

In the first example you see an object (file+meta) "ABCDEFGHI" stored in a classic N-copies (3 in this case) object strategy. If you look close you'll see that the entire file is written to 1 disk of 1 storage node and could be distributed to 1 disk per copy, whether it is in the same node for disk failover or on another node for disk+node failover.

In this second example I'll show you how Amplidata stores the object. The file is divided in CHUNKS with a fixed size. These chunks are then stripped down again in an amount of parts, called SHARDS. Shards are 1/4096th piece of the chunk + ECC. In this specific case we have an object "ABCDEFGHIJKLMNOPQR". The object is split in chunks of xMB which are divided in 4096 shards and then distributed over for example 9 disks. The orange part is the overhead for ECC. I'll explain failover in the next part.

Failover Basics/Mathematics:
Actually it is pretty simple. A protection policy has 3 basic characters: the size of the chunks, the amount of disks for distribution and the amount of failures. The higher your failures/disks ratio, the bigger your ECC part will get.

Example: We have a file of 1228MB and a Policy of 32MB chunks / 16 disks / 4 failures. This means we will get 38 chunks of 32MB (and 1 of 12MB at the end), which are always divided in 4096 shards. The bigger the chunks (ex. 64MB), the bigger the shards (would be 16KB). Now it is time to do some failover. We asked for 4 failures on a total of 16 disks so we will get 6720 blocks containing data + ECC coding. As you see in the example on the right we never just save parts of the data, only calculated results after processing. Now comes the tricky part: when reading the data we need ANY 12 out of 16 disks to read from just because of that distributed calculation model.

So we only have 25% overhead in this example. Well, there is an additional overhead for ECC coding itself (20%) which would end up around 150% for 4 failures but if you now look back at the former example we needed 300% RAW storage for only 2 failures. The best part of it is when you design your policies, the GUI will tell you the overhead you are creating. I saw a real life example where we tried a 20/5 policy and it told us that we needed exact 1,52 netto/raw ratio (on a 20/10 that was 2,28). Sizing: chunks can have several preset sizes between 256kB and 64MB and the default is set to 32MB. The max amount of disks in a policy is now 20.

note: you see I used chunks & shards and in the same lines you see super/message/check blocks. The first 2 are universal terms, the last are Amplidata's terms.

Benefits:
Besides a huge difference in overhead the real storage people out there will already have seen another big benefit: THROUGPUT! You are writing/reading to/from a lot of spindles on as much as possible nodes. Not just 1 spindle on 1 node at the time.
This is what Howard came to do in Belgium. He ran some benchmarks against the whole thing and his results were amazing (in comparison to others). Read more about the review in the link at the bottom of this page.

Architecture:
I knew before I went to Amplidate HQ that they were selling the products with their own hardware. At first I thought this was really wrong. If you are a startup you should be focussing on the software. However ... this is what they showed me onsite:

Do you know ANYONE out there having 12x3,5" disks in 1U? I mean, this is freaking 36TB of RAW data (48TB when 4TB is fully supported). This is 1,3PB today (1.5 PB soon) of raw data in a single 42U rack (leave some space for compute and switches).

And then there is the physical scale-out architecture. The system has compute nodes and storage nodes. The one I saw that Howard tested was a 3 compute + 24 storage node configuration. There is no specific ratio of compute per storage and there is no theoretical limit. The compute nodes have 2x10GbE uplinks and 2x10GbE downlinks to the switches. The storage nodes all have 2x1GbE link to the switches. The communication between the compute and storage nodes is an Amplidata proprietary IP-protocol.

The compute nodes handle the breaking objects down in chunks/shards and distributing to the storage nodes. The storage nodes however are not JBODs! They handle some stuff on their own like background scrubbing. Scrubbing is a technique where you look at all the physical data and confirm if it is still good or not and rewrite to other blocks if not.

My Take-Aways:
As usual I share some of my personal insights here.

We already know that when you have A LOT of data, block storage is not going to be of any help. Even filesystem storage has its limitations. Object storage is not new but now it really gets some traction. Imagine those billions of facebook pictures, youtube movies, MRI scans with very high resolutions, ... From what I have seen up until today, there is no "best solution" build yet. And if it has been build, it might just have been a purpose build solution that is not good enough for a broader use.
I like this one. They have a few years of lag on some others in the industry but they played it smarter from a technology standpoint. I just hope it's enough to be disruptive. Don't get me wrong here; this is not R&D only. This is already Generation 2 and runs in production in pretty big environments.
Amplidata has a multi-rack, multi-site approach that I didn't cover yet (blog was already long enough). It gives way less overhead than the competition but there are still some downsides I want them to tackle first. One of those things is that even if you have enough chards in site A to recover the object (data+ECC) it will retrieve all chards available from other sites to just read the object. This would not affect the performance to much but it will eat your bandwith unnecessary.
Amplidata should be in the next Storage TechFieldDay! The delegates would love it!

Links:

Corporate blog: LINK

DeepStorageNet review (by Howard Marks) - LINK (will come when available)
Press release announcing new CEO + $6M new venture capital.
Press release announcing the review by Howard Marks (will come when available)

What is Scale-Out?

2012-10-11T23:34:00.000+02:00

A little while ago I did 2 posts on Nexenta. In the first one I gave some basic insights in ZFS as a file-system and in the second I tried to give a glimps of how this can be scaled up/out. In those solutions I also mentioned AoE (Ata over Ethernet), a technology developped by CORAID that presents disks(cabinets) raw over Ethernet without any form of SAN. I can truly understand if this doesn't ring a bell, it didn't for me too the first I heard about this at Storage Field Day 1 (SanJose). This post goes a little deeper on AoE but also on what is now actually real Scale Out.

What is AoE?

AoE is a purpose built protocol built on top of Ethernet. AoE is a connectionless network protocol that does not rely on a session or transport layer. Coraid’s AoE driver recognizes every port and will use all in parallel. When an AoE target presents itself to the initiator, the driver will sense multiple unicast lanes between the server and storage without the need for a MPIO driver. When the OS drops an IO request (read or write) onto AoE, the driver will segment the IO into multiple “max MTU” Ethernet frames. These frames will be sent across every available HBA port as unicast AoE datagrams to the AoE target. Each of these datagrams must be acknowledged (with the unique correlation tag) prior to acknowledging the IO to the operating system, ensuring reliable delivery. By using multiple unicast lanes, Coraid’s implementation of AoE is able to achieve much higher throughput and lower latency (response times).

Source: http://www.coraid.com/solutions/technology

In the end this is only a storage backend. You still need some kind of NAS/SAN frontend to present these physical boxes to the servers/clients.

Coraid ZX NAS
A few weeks ago CORAID announced their own NAS headers: ZX-Series. As they already partner with Nexenta for scale-out NAS solutions, I assumed this was a logical next step for both. Apparently it is not. I took me a few tweets to finally get the following answer: "Coraid ZX-Series is based on Oracle Solaris 11 and RSF-1 from http://www.high-availability.com". So why did they make this decision? John Gilmartin, VP Product Management clarified this for me:

Coraid is a big fan of Nexenta, and the combined solution provides powerful differentiation over typical white box deployments. We are sister companies in the Menlo Ventures fund, and continue to partner with them in accounts worldwide as part of our OpenFS strategy...In terms of our ZX solution, there were a number of business and technical factors that went into the selection. Oracle Solaris is a great OS which continues to evolve, and it includes all of the powerful features of ZFS that we needed.

Scale-out NAS
But then a few days ago I came across this post that stated the ZX would not be able to cluster. If you read the post it is a bit fallacious. What they really meant was that for now there is no scale-out clustering. As far as I know, a high availability cluster of 2 nodes is also a cluster but maybe that's a pet peeve of the writer or just a way of drawing some more attention (it worked for me). But where he had a point is that a 2 node cluster (as the Nexenta for example right now also) is not a Scale OUT cluster. Scale out is many-2-many.

Real Scale out NAS
For real scale out you have to look for example to NFSv4.1. In that version pNFS has been added. pNFS stands for Parallel NFS and is an open-standard (!= open-source) supported for many years by Panasas. pNFS has been an answer for quit a while to HPC (high performance computing) as it addresses the problems a regular file-system has when addressing al the IO through a single file header. Instead of reading the IO's through the storage header, the namespace servers will direct the clients to the repositories:

Source; http://www.panasas.com/products/pnfs-overview

My TakeAways

I truly believe file-based/object-based storage will drive our storage choices more than block-based solutions. More on this in a next blogpost on a Belgian startup in the object storage: "Amplidata".
iSCSI has made a huge step-up in Enterprise storage but one of the downsides to most of the iSCSI products is that after a while multipathing can become struggle. One of my pet products EqualLogic for example can't handle more than 1048 iSCSI connections per storage group. I'm sure new FW would push that number but not faster than we are scaling now. ANd I never really liked FC for its complexity. With AoE you don't have these issues so much.
CORAID AoE is also compatible with Windows Server. I haven't given this a really deep thought yet but I can imagine the new storage capabilities of Server 2012 can leverage from this backend when it comes to scale (Windows SOFS supports 64 nodes).

OTHER LINKS:
Ray Luchessi: Coraid, first thoughts
Aidan Finn: Scale Out File Server in Windows Server 2012
Martin Glasborow: Clustered NAS versus traditional NAS

VMworld Europe - start preparing

2012-09-08T09:49:00.001+02:00

Espana Por Favor!

This post will be updated as new information comes in. Social gatherings, tips and tricks on local places, Veeam announcements (off course), Spousetivities!, ...

Spousetivities:
let me start with a big cheer for Crystal Lowe (@Crystal_Lowe / @Spousetivities). She does a hell of a job organising this thing. We suspect that she even runs a bigger show than her husband Scott Lowe. For the first time at this event I am very proud that my wife (@LynnStruylaert) will join me and with that Spousetivities.

Planned activities:

Dali Tour - Tuesday, October 9
Barcelona Tour - Wednesday, October 10
Montserrat and Cava Tour - Thursday, October 11

Website Spousetivities: LINK
Registration landingpage: LINK

VMUG Party:
As usual #Veeam tries to support the community as much as we can. Therefore we are very proud to be the official sponsor of the VMUG opening party.

Grab some great giveaways
Win a free VMUG Advantage subscription
Talk with the VMUG Board of Directors and local VMUG leaders
Get connected with your local community and request new local VMUGs
Provide some on-the-ground reporting. We'll have a live Twitter feed in our booth - make sure to use the #VMUG hashtag to participate

VMUG Party: LINK

Veeam Party!:
Another BANG! The most beloved vendor party of VMworld US could not stay behind in Barcelona. On Octobre 9th, we are taking you to Club Shôko.

Veeam landingpage for VMworld EMEA: LINK
Veeam Party - RSVP: LINK

#vBrownBags:
#vBrownBags is an independent community of technology specialists that share knowledge. If you ever want to go after VCAP certification for example, you definitely need to subscribe to the #vBrownBags podcast.
I am very proud that #Veeam is sponsoring #vBrownBag to come to EMEA. We did it before at VMworld US but this one will also give the European people some possibilities to share knowledge. And guess who has scheduled a session on Wednesday octobre 10th at 14.45PM? Yours truly!

#vBrownBag landing page: LINK
Podcast on iTunes: LINK

Google Map with all those important places (to me):
(tip: zoom out once to get an overview)

Other interesting links:
Official VMworld Barcelona landing page
Fira Barcelona - VMworld venue
Social Gatherings landing page

Touristic tips:

every foodlover should at least have visited la Boqueria once, the covered food market on the side of the Ramblas.
If you love Gaudi and you want to enjoy his weird architectural inventions I have a few for you. For a stroll in the park you should visit Park Guell. You could do the entire park in an hour or 2 without being in a hurry. Ideal place to go if you are leaving a day later to cool down.
I know it is the most cliche building in town but if you have never visited the Sagrada Familia, you will regret yourself going home without visiting it. Make sure you are smart when you go because you will encounter long waiting lines. Interesting fact: this church is in construction for 130 years now. This is because it can only be funded with money the community raises (so no sponsorships or subsidies).
Want to have a birds eye over the town and the harbor? Take the Port Cable Car (teleferic) from the harbor to Mount Montjuic.

The Gospel of Hans

2012-07-25T22:27:00.000+02:00

A technical or technology evangelist is a person who attempts to build a critical mass of support for a given technology in order to establish it as a technical standard in a market that is subject to network effects.

It's been a nice ride these last few weeks. Less than a month ago I publicly announced resigning my current vacation as a Solutions Team Lead and I am already looking forward to this new opportunity.

VEEAM speed
One of the first things I came across in my search for this new gig was "VEEAM speed". I know some local people like @bertarnauts and @d_steven that already planted an egg in my brain how much they loved working at VEEAM and off course @rickvanover did some effort too. But when I said I'd be available for new opportunities they took this to a whole new level. Soon I was in contact with Doug Hazelman who was looking for an evangelist to cover the entire EMEA market. Because Doug was in Europe he wanted to do a face2face interview. So I made a trip to Zurich two weeks ago. Apparently it was some kind of Executives quarterly meeting so I got interviewed not only by Doug but also by Ratmir Timashev, the CEO of VEEAM. At first this felt really awkward but we managed to have a very interesting conversation and I very much appreciate the time he took to talk to me.

Software? For real?
Yep, software. As you all know by now I was the iron man, focused on datacenter hardware, specifically storage. Although I have been talking to some storage vendors those jobs were either too technical (I don't dream in syntax) or too local. I love talking about storage but the most fun part in that line is the talking. So being an evangelist for a software company I like tops being an engineer for a storage company (I like) any day.
Why did I agree on the VEEAM offer? In the interview in Zurich Ratmir Timashev told me I could ask him some questions too. He told me before that they have been able to double the figures year after year and would be keep on doing that and becoming a billion dollar company one day. I asked him which competitor would be able to get in his way of getting there. If there is one to fear, who would that be? He really had to think a few seconds and honestly answered me: none! And I believed him instantly.

The Dark Side
Yep, it has happened. Everyone knows how much I loved being an independent and it took me a while before accepting that going to the dark side was in fact the only viable opportunity at this point. I promise that I will keep following enterprise IT and storage in specific on this blog and that I will always try to keep my unbiased glasses on. Please help me when I lose that! I will miss the invites that I might be missing as an independent blogger (i.e. Storage TechFieldDay). We might catch up on the same events but I'll probably be wearing a green shirt then. So instead of the DARK side I'll make it my GREEN side!

So maybe that HULK theme is still going to work out for me someday :-)

HP Virtual Connect and Flat SAN [Infosmack podcast]

2012-07-23T18:41:00.000+02:00

Infosmack has always been one of the most respected podcasts within Enterprise IT. Therefore it was an honor for me when Nigel Poulton asked me to co-host an episode on HP Virtual Connect and their new-kid-on-the-block "Flat SAN". We were joined by Alex Kramer, Sr. Software Engineer and Architect at HP.

Podcast

Your browser does not support the audio element.

DON'T MISS AN EPISODE - SUBSCRIBE FOR FREE!

Subscribe through iTunes

Subscribe through Google

Virtual Connect
Virtual Connect is a technology HP introduced 5 years ago and is their first solution towards the virtualization of the network environment. With the integration of 10GbE and Converged Network Adapters (CNA), they were able to split a single port into 4 different physical PCIe channels. On top of that you can choose whether these channels were to handle Ethernet or Fibre Channel protocols. I'll show this in a common use case:
Before: BL460c, 6x1GbE and 2x4GB FC

After: BL460c (G6 and up), 6x1GbE and 2x4GB FC

You have the same result, but with 6 interconnect switches less and every blade has 2 extra free PCIe slots for more connectivity (example SAS JBODs) or local SSDs (example FusionIO).

Link: HP Virtual Connect for Dummies
note: I dislike the fact that once again a vendor requires you to register to read a white-paper. If you want us to read us as much as possible, please make it as easy as possible (=one-click and anonymous)!

Products
Be aware of the difference between Flex10 and FlexFabric interconnects and CNA's! With Flex10 you can only use IP protocols, for FC you will need FlexFabric. Both are under the Virtual Connect umbrella.
The FlexFabric CNA's are based on the Emulex OneConnect 10GbE FCoE (554FLB - LOM / 554M - Mezzanine)
The Flex10 CNA's are bassed on either Emulex OneConnect 10GbE (552M / no LOM) or BroadCom Nextreme (530FLB - LOM / 530M - Mezzanine)
There is also a third adapter which only does 2x10GbE which is based on Intel 82599 10GbE Controller (560FLB - LOM / no mezzanine).
Note: I only used the Gen8 product numbers here but G6 and G7 should do most of the same technology.

Flat SAN
At HP Discover they announced the next generation of VirtualConnect and introduced the Flat SAN. what is a Flat SAN? It's an architecture where the Storage Array is directly connected to the servers without the need of Fibre Channel SAN (SAN = Storage Area Network). Eliminating the need of ToR (top-of-rack) Fibre Channel switches and thus reducing a lot of extra costs in the datacenter, simplifying physical components and management.

If we have a look at the biggest 3PAR (V800) who has a maximum host port connectivity of 192 ports, you can direct attach up to 48 blade enclosures (4 connections per enclosure) to a single Storage System, handling no less than 768 blade servers.

So how did they make this possible today and were not able to do that yesterday? Formerly the VirtualConnect switches used only NPIV for all ports and have the SAN zoning done by the ToR switches. Now the VirtualConnect switches change their ports in native Fibre Channel fabric ports when they see a 3PAR connected. The SAN zoning/login is now done in the VC switches.

My TakeAways
I know they did this specifically for the scaled architectures but the sooner HP is able to make this model independent (so including P6000 EVA and P2000 MSA) the sooner it will get integrated all over the place.
And PLEASE don't stop there! make sure you do every possible effort to integrate ALL fibre channel arrays, regardless of the vendor. This will really give you the benefits you need in the datacenter (small and big).

Watch Calvin Zito (@HPstorageGuy) and Brad Parks (@HPBradParks) introducing the Flat SAN at HP Discover 2012

LINK: Calvin also did a podcast on ATSB (Around The Storage Block) with Alex Kramer. Here is their landing page with some extra links.

Nexenta scale and cluster

2012-07-02T00:02:00.001+02:00

Last month I blogged on the basics of ZFS and Nexenta. Apparently everyone liked the tiering superheroes. In fact I started with "the Flash" for the SSDs but I figured no one beats the Flash and DRAM is flash too. So I picked the versatile Superman for the L2ARC/ZIL part of the job (could have gone Green Lantern here too). So far for the fun part. Hope I got you interested in the ZFS file system, now I'll get a bit more serious on the Nexenta failover techniques.

1) Disk management:
The first part of failover management is still plain old ZFS. It's the way disks work together, how I/O is handled and how data corruption is avoided.

Storage Pool: a set of (equally configured) vDEVs
vDEV: a set of disks with a certain protection level
RAIDZ: first form of physical disks protection level. Examples:
RAIDZ-5: 4 disks data, 1 disk parity
RAIDZ2-6: 4 disks data, 2 disks parity
RAIDZ3-9: 6 disks data, 3 disks parity
Mirrored Disks: second form of physical disks protection level. Here too you have multiple choices of failover (dual-mirror, triple-mirror, ...)

Stripes: writes are striped across all members of a vDEV, therefore the speed/bandwidth will be truncated by the slowest member. To add bandwith/IOPS, add more vDEVs. But this is for parallel IO streams! One IO stream = 1 vDEV.

RAID5 write-hole: ZFS is protected against the RAID5 write-hole because it does copy-on-write instead of read-modify-write. If in case of a power failure some blocks are not completely acknowledged (incl parity-check), nothing would have happened.

Physical disk-enclosure failover: if you design smart, you could even design for disk-enclosure failover by aligning the vDEV physical disks vertically (1 per enclosure)

2) System Scale:
The next step is in-system scale. It's not really high availability but an equally important step in the story going forward. Because you benefit a lot by a bigger bandwith with ZFS you'd want to "scale-out" first inside your system and then go to "scale-up". Here you find a simplified example of that.

So instead of daisy chaining the 3 first nodes and then adding a second and 3rd loop I did it the other way around so you'd have 3 loops of 6Gbps SAS connectivity from the beginning.
Remark: this is a theoretical model. I would have to doublecheck with the solutions team before designing this for a customer.

2) High Availability Cluster:
As far as you have seen by now I only used 1 controller node and some scale-up possibilities. There is still 1 pretty big single-point-of-failure: the 1 controller node. Nexenta took care of this by adding cluster technology. The basic is: any vDEV is owned by 1 controller but can be accessed by multiple controllers for failover. There are 2 failover parameters used: the first is a disk heartbeat, the other a network heartbeat.

How does this work?
When you add a node to the cluster it needs to be connected to the same jbods and it will be assigned heartbeat disks and a heartbeat ethernet link. When both fail on node1 the second node will notice this and will put an ownership byte on the heartbeat disks (only heartbeat disks since 3.1.3) will take around 20 seconds and will be assigned ownership on the pool and will import the pool. when node 1 comes back online you can switch the vDEVs you want back to node 1. It will not do that on it's own. It will however take control again in case Node2 would fail.

First thing that popped in my mind is this: where is the IO ack? As you have seen in the previous post the ACK is in the ZIL. So you'll have to put the ZIL on the JBODs to be able to safely failover at any time. (If you don't get this, stop reading. You're not supposed to hear this)'. Although a Nexenta Cluster would have only 2 nodes in a classic HA Cluster they already scale up to 8 nodes if necessary.

Now think crazy, take a whiteboard and some markers 'cause you're going for a wild ride: there is no boundaries to the connectivity of disks & controllers as long as you can physically manage to connect them. So if you were adding SAS switches or even Ethernet switches (AoE like Coraid) you could en up with real scale-out AND scale-up into petabytes within no-time. I have made one example below:

Right now, you are looking at a 4 controller 500TB (4x120x900GB 10k) scale-out NAS you have build on your own. I did this on DELL hardware because that is something I know best (could have went HP too off course). Only downside here is that little fact that the ZIL SSDs are not on the CPU side of the HBA. Today Nexenta advices the use of STEC ZeusRAM DRAM SSDs. I'll have to add them somewhere on the JBODs but as I don't have best-practices experience I'll keep this one on the theoretical side.

3) Multisite Cluster:
So what happens nowadays when you have multiple sites? Every customer asks you to do Active/Active SAN over 1000 miles distance. You know what? Ain't going to happen :-)
No, seriously. Anyone still reading knows by know that we have to educate the customer and his/her demands. So what are the possibilities Nexenta has to offer (technically possible whitin acceptable latency!)

Autosync (async) replication this replicates the metadata also so it preserves snapshots (preferred choice #1)
Block level Sync replication with autocdp (CDP = Continous Data Protection). This is sync but please be aware of the latency here.
Mirror across 2 sites (advanced set-up!). This is a true metro-cluster. I will not elaborate too much here as it is only custom made architecture.

This truly is Enterprise Storage for Everyone. I can assure you that buying this building blocks and adding Nexenta on top is way cheaper than any solution you are going to buy from those big vendors. And on top none of those would have all the features Nexenta has builtin. Think 128bit filesystem, VAAI, NFS4, iSCSI/FC/FCoE/FCoTR, AoE, SMB3 (to come), ... Now take this one step further and really buy your stuff off the shelve ... $$$ > $$ > $!

MY TAKE-AWAYs:
There are still a few key-points I'd love to see them do differently or at least give you the chance to do so. Example: the fail-over is fine but there is no automatic fail-back. Add an auto-balancing of the vDEVs ownership over all the controllers and you would have a sweet thing out there. Another point in that "peer2peer ownership" is that you'd have to do quite some manual work adding/removing nodes. Like when you want to change on generations of controllers. Wouldn't it be nice to add 2 new nodes en just ask the software to remove 2 older nodes from the cluster in 1 command?

LINKS:
Nexenta HCL: start designing HERE!
Nexenta LIST prices TIP: buying from a reseller is usually cheaper and if you pick the right one, you'll get some great consultancy along.
Nexenta Tech Briefs

so I resigned?

2012-06-23T11:08:00.002+02:00

Let me first start by saying that I do not blame my current employer that this didn't work out on a long term. I did get a lot of opportunities that helped me being where I am right now but my day2day work turned out not to cope with my personal expectations.

A second, equally important reason, is that I recently discovered that doing a job just because you can is not important enough anymore. I would like to go on a quest for the job I really love and that fits my personality more.

I did not take this decision with any new directions already in mind so I will be looking at all opportunities the next few weeks. Feel free to contact me if you would like to receive my resume. Remark: if I would be looking for another presales or system engineer job I'd better stay where I am :-)

Stay tuned!

Small is the new Big [drobo announcement]

2012-06-21T08:54:00.001+02:00

It's time to give another one of the Storage TechFieldDay 1 (#SFD1) vendors its turn to get some extra spotlight. You might be wondering what a company like Drobo actually does on an event that covers Enterprise Storage solutions. Don't they make those little boxes to put commodity disks in for your desktop? We wondered too.

Apparently Drobo has done a lot of work getting up the stack. When we were in SanJose they showed us the B2100i, a 12 bay iSCSI SAN based on the disks you prefer (as in "buy of the shelf"). They still miss quite some features to call it Enterprise Storage but they come in close with auto-tiering and support for up to 3 SSDs which are used for transactional data. In fact Drobo knows their market is not the Tier1 applications. They are here for Tier2-3, LAB or HOBO (HomeOffice - BranchOffice). The punchline here is "my OTHER san is a DROBO". I like that insight.

So, you heard me saying Auto-Tiering and SSDs for transactional data. Mix that with TechFieldDay, some greybeards (yes Howard, that's you) and a vendor that actually read the memo on how to present to techies. The result:
I'm the silent guy (in this session) in the blue shirt slightly hammered by 3 days of no sleep as a result of being on the wrong side of the ocean.

SO, what's new today?
Today Drobo announced the next step for the smallest devices. And when they say small, they mean small. Please welcome the Drobo mini :

What is that?

A ruggedized metal carbon fiber case that can hold 4 standard 2.5" drives AND an mSATA SSD. I honestly think this is a very sexy product and will definitly try to get my hands on one as soon as possible. Drobo announced they should come in around $650 as of somewhere throughout July. Have you already thought of the possibilities of carrying your own lab environment in your backpak? No more hustle to figure out why the VPN doesn't work or why your PC's local storage doesn't handle all those virtuals ...

Besides the Drobo mini , they also announced the Drobo 5D . It has all the same capabilities as the Mini but with 5 x standard 3.5" drives AND the optional mSATA SSD. Another word from the product managers was that the 5D would eventually replace the standard Drobo and the Drobo S.

Advantages:

4 (or 5) drives + mSATA SSD
auto-tiering (out of the box so not tunable!)
another sexy thing not yet mentioned: the Drobo 5D and the Mini support not only USB3 but also the THUNDERBOLT which operates at a maximum of 10Gb/s and is able to be daisy chained up to 6 members. Therefore each Drobo has 2 of these connections.
LEDs that show status and usage
up to 5x performance of current models

LINKS:
Press Release
Drobo Mini Datasheet
Drobo 5D Datasheet
#awesome read on MyDigitalSSD mSATA by @theSSDreview

disclaimer: all transportation, catering and stay on Storage TechFieldDay was taken care off by the sponsors through the organization of GestaltIT

MSSQL must really love Oracle (pun intended)

2012-06-06T00:38:00.001+02:00

You remember everyone ranting on the Oracle licensing schemes? You remember everyone ranting on VMware last year when they announced the vTAX licensing scheme?

Today I was reading through the new Microsoft SQL Server 2012 Licensing scheme and got struck by lightning (not in any positive way). I'll elaborate on what happened/changed and I sincerely hope anyone will correct me soon!

I will not go through all possible differences in licensing schemes and focus on 2 examples:
1) a single commonly configured physical SQL server Enterprise with 2x6 cores
2) a single commonly configured virtualised SQL server Enterprise with 6 vCPUs on a 2x6 core hypervisor.

SQL Server 2008 Licensing:
Under SQL Server 2008 a physical server was licensed per physical socket.

Microsoft offers a Per Processor licensing model to help alleviate complexity. When licensing SQL Server software under the Per Processor model, you do not need to purchase additional CALs; it includes access for an unlimited number of users or devices to connect from either inside or outside the firewall. Per Processor Licenses for SQL Server 2008 R2 are available for
Datacenter, Enterprise, Standard, Workgroup, and Web editions.

Notes:• A Per Processor License is required for each processor installed on each operating system environment (OSE) runningSQL Server or any of its components (for example, Analysis Services).• For SQL Server running in physical operating system environments (POSEs), you must license all physical processors.• Per Processor License costs are the same regardless of number of cores in the processor.

If you figure the cost of 1 SQL Server License is around €25.000 we would have a total license cost of €50.000 for our physical SQL Server Enterprise Edition because we had 2 physical sockets.

Licensing for Virtualization Under the Per Processor Model

The number of operating system environments (OSEs) in which you may run instances of SQL Server 2008 R2 under the Per Processor model depends upon the edition you license and whether or not you license all of the physical processors with a Per Processor License.

Licensing All Physical Processors
If you license all of the physical processors on the server (one license per physical processor), you may run unlimited instances of the SQL Server software in the following number of OSEs (either physical or virtual):
• SQL Server Datacenter = unlimited
• SQL Server Enterprise = 4 OSE's

Licensing a Portion of the Physical Processors
If you choose not to license all of the physical processors, you will need to know the number of virtual processors supporting each virtual OSE (data point A) and the number of cores per physical processor/socket (data point B). Typically, each virtual processor is the equivalent of one core

In our case we would chose only to license the Portion of the physical processors. As we wanted 6 vCPUs in a 2x6 core hypervisor this would have a total license cost of €25.000 per virtualised SQL server. Important restriction: you cannot move this VM over a cluster of more than 4 hypervisors.

SQL Server 2012 Licensing:

Under SQL Server 2012 Microsoft changed the licensing model from per socket to per core. This changes A LOT!

SQL Server Licensing Options:
SQL Server 2012 will continue to offer two licensing options - one that is based on computing power, and one that is based on users or devices. In the computing power-based model, however, the way we measure power will shift from processors to cores.
• Enterprise Edition (EE) will be licensed based on compute capacity measured in cores.
• Datacenter Edition is being retires with all capabilities now available in Enterprise
• The Enterprise (and the Standard) Edition of SQL Server 2012 will both be available under core-based licenses. Core-based licenses will be sold in two-core packs.
• To license a physical server properly, you must license all the cores in the server with a minimum of 4 core licenses

A two-core license costs around € 20.000 per piece (incl SA for 2 year). So our physical server with 2x6 cores would have a total license cost of €120.000 (6 x 2-core license).

Virtualization Licensing - Cloud Optimized
... There will be two primary virtualization licensing options in SQL Server 2012: The ability to license individual virtual machines and the ability to license for for maximum virtualization in highly virtualized and private cloud environments.
Individual Virtual Machines:
• To license a VM with core licenses, customers can simply buy a core license for each virtual core allocated to the virtual machine (minimum of 4 core licenses per VM).
• Each licensed VM that is covered by Software Assurance (SA) can be moved frequently within a server farm or to a third party hoster or cloud services provider.

Lets do the math again on our 6vCPU virtualised SQL Server: 3 x two-core license would have a total license cost of €60.000.

Conclusion:

I think I already gave you my conclusion by writing this piece. And I really think I did not exaggerate. Actually I wanted to do the same equation with a DELL R820 4x10 core server but I guess you can do the maths on your own by now.

Extra tip:

Maybe this is the point you might want to reread my post of last December on EnterpriseDB. They have a really nice licensing scheme (subscription scheme) that if I told you the cost you wouldn't even bother redoing all the maths above.

LINKS:
Micosoft SQL Server 2008R2 Licensing datasheet
Microsoft SQL Server 2012 Licensing datasheet
Enterprise DB

EqualLogic connectivity TIP!

2012-06-04T23:40:00.001+02:00

Today I accidently stumbled upon something I didn't know about the connectivity in the DELL Equallogic PSx1xx models (PS4100, PS4110, PS6100, PS6110). Let me start with the basics and work my way through to the interesting point.

EqualLogic controller basics:
Each DELL EqualLogic array member already has 2 controllers (you could buy a single controller version but why would you?). Each controller has off course CPU, RAM and Connectivity. The 2 controllers are in a Active-Standby mode. This means:
The CPU and RAM of both controllers is used but only the connections of the active controller.

Basically this means that at any given time the active controller (A) can fail and the only thing needed on controller B is activating the adjacent ports with the same IP addresses. EqualLogic also uses this procedure when upgrading firmware. First the firmware of the "standby" model is upgraded, then the connectivity is changed so Controller B becomes active and anfterwards controller A is upgraded too. The advantage here is that you only need 1 controller failover for a 2 controller upgrade.

Vertical NIC failover:
With the PSx1xx models DELL Equallogic took the failover level one step further. If a network port would fail, the adjecant port on the other controller would take over. As that controller is in "standby" mode this is not really an issue. This prevents a failover procedure, rebooting the controllers.

The Challenge:
Today I was designing the difference between a PS6100XV solution and a PS6110XV solution for a customer. Looking at the design I noticed a SPOF (Single Point of Failure): the switches! What happens if not the controller fails but the switch. In a normal configuration with a PS4100 or a PS6100 you would have 2x1GbE or 4x1Gbe per controller, equally devided between 2 switches. This would not give any issues. But on a 10GbE controller you only have 1x10GbE per controller, hence only 1 active connection.

It appears now that the vertical failover mechanism kicks in here too. The trigger is not the physical NIC failure but the connectivity failure. So whether the NIC is down or the port on the switch goes down or the entire switch goes down ... THERE IS NO SPOF!

AHA Erlebnis*:
If you look back at the PS6100 model you'll find an opportunity here to go further than we went before on failover levels: crossconnect the vertical failover ports to different switches! This would always provide you with 4x1GbE connectivity even if one of the switches fails!

* Aha-Erlebnis: sudden insight / Eureka moment

VMUGbe #17 - bloggers edition

2012-06-04T21:22:00.000+02:00

Last friday (june 1st) VMUGbe went big! For the first time we had +150 people attending. #VMUGbe went through some changes lately: Erik Schils, VMUG leader of Belgium finally joined twitter (VMUGbe) and we became part of the VMUG.com . For the ones that don't know how this works: every VMUG is is free to chose between staying independent (like VMUGnl still is!) or joining the global .COM community. The (dis)advantages are obvious. For a smaller community like ours being part of a bigger organisation is definitely going to help us grow faster. For VMUGs that are big enough on their own, the rules & regulations of a larger community might sometimes be a constraint.

How could this event become so big? Invite the (Dutch vMaffia) bloggers and you get all the attendees for free :-) See the lists of sessions for yourself and imagine you could have been there. You won't have any second thought I assure you.

RAID types and alignment impact on VMware workloads - Erik Zandboer
Workshop “Designing a cloud infrastructure” - Duncan Epping & Lieven D'Hoore
“Beer and PowerCLI: you can't live without them” - Luc Dekens
“HyperV v3.0 vs VMware vSphere 5” - Alex Muetstege & Erik Scholten (VMGuru)
“Troubleshooting VDI” – Sven Huisman
“VMware Health Check” - Gabrie van Zanten
“Busting a few myths” - Eric Sloof
“Mobile devices and Cloud. Bring 'em on” - Albert Kramer (Trend Micro)*
“VCE” – Cisco, EMC & Trend Micro*
“Storage DRS” - Frank Denneman
“Unleash the IAAS Cloud: About VMware vCloud Director and more...” - Viktor van den Berg
“vCenter Orchestrator” - Arnim Van Lieshout

The two greyed out sessions were sponsored. Most of you know my strong opinion on sponsored sessions: make them interesting or don't make them at all! This VMUG was a very long day full of interesting sessions and these two just didn't make it to the requested level. I have two very short tips for any vendor/reseller who will present at a User Group event:

Don't tell us the problem! We already know the problem.
Leave the marketing/sales people behind their desks and bring in the architects! The people in front of you are techies, convince them and the sales people will get their own invitations automatically.

Got you attention? Read this full blog post of the Tech Field Days with all do's and don'ts of presenting to technical people.

As you might have noticed we only had 2 Belgian bloggers (Lieven D'Hoore & Luc Dekens). Normally we should have had another session by Gert Van Gorp but he had to call off being sick. Some of the bloggers said I should be ashamed I was not on stage. So I already promised Erik I will prepare a session for next VMUG. Maybe I could do something around new storage products for that fit well in the VDI space. More on that later without a doubt.

Vendor spoof: at the end of the sessions "a" vendor gave us a spoof on their release of a new product that came out today. Although I love the guys bringing it I couldn't sell out cheap here, blindly repeating their marketing message. Next time if I have some more time to work on the product in advance I might do a specific product special. Sorry for the missed Wodka bottle opportunity :-)

I detest what you write ...

2012-06-04T18:31:00.000+02:00

"je déteste ce que vous écrivez, mais je donnerai ma vie pour que vous puissiez continuer à écrire" - Voltaire "I detest what you write but I'd give my life for you to be able to keep on writing"

A certain discussion on twitter a few weeks ago forced me into this post that maybe each "independent" blogger has at one point in time. It's about BIAS.

For starters let us be clear: every blogger is biased through his experience. If you work/have worked for a vendor, you are biased by your experience at that vendor. If you are a reseller you will be biased by the companies of whom you sell products. If you are a channel partner / systems integrator you most definitly will be biased towards the solutions you provide.

Personal Disclaimer: I work as Solutions Team Lead for a VAR (Value Added Reseller). Our primary partners are DELL, HP, VMware, Microsoft and Cisco. Within that function I do the presales part of servers & storage and vendor management for HP & DELL. Being online at the different communities from both vendors, twitter, G+ and this blog I sometimes get invited to their venues as "independent" blogger.

What is acceptable BIAS for an independent blogger?
As I have no IBM, NetApp or EMC products in my portfolio I will not give criticism on the technical part of those products. What I can do is questioning if the value prop they are given is no marketing BS (ex. EMC Project Lightning or NetApp onesizefitsall). People who follow me closely know that I will stand corrected when necessary but will not apologize if you feel offended (as this cannot have been intentionally). Even more: I look forward every day to be corrected and learn more from all you smart people out there.

What is NOT acceptable?
As we all agree, spreading FUD (Fear, Uncertainty, Doubt) intentionally is one of the worst things we can do, independent or not. What is the margin between a personal opinion and spreading FUD? This one is thin and I'll keep it up to each and everyone of you to evaluate your posts on that. What I really have a strong opinion against is FUD from one vendor to another. If you write blogposts on the website of a vendor (hp.com, dell.com, emc.com, netapp.com, ...) your post will be read by the entire world as a message from the company. Customers, analysts and press will copy your point of view as being corporate. Let them handle the criticism and focus on the strengths of you company. If you can't let it go, get a blog in your own personal name and write your opinion.

Interesting link: Stephen Foskett wrote about how to preserve our credibility as independent bloggers.

Nexenta & ZFS basics

2012-05-25T00:39:00.003+02:00

Today I was invited at the OpenStorageSummit EMEA in Amsterdam. They could have just called it Nexenta Summit but hey, who cares? It was a really nice summit. Not to big but everyone you could possibly imagine was there. It was great meeting for example CEO Evan Powell and VP Marketing Bill Roth amongst alot of local Nexenta sales people as global techies (Theron Conrey and Andy Bennett).

So what the heck is Nexenta doing? No more or less than ZFS on steroids:
"...The core storage platform, NexentaStor™, is a software-only storage operating system based on Linux and a storage-optimized file system based on the OpenSolaris / Open Storage ZFS file system. With this unique combination of open source codebases, NexentaStor delivers the ease-of-use and developer-friendly aspects of Linux in tandem with the power and scalability of the revolutionary ZFS file system..."

Yes, software-only! This means you literally can choose whatever commodity hardware underneath that you want to. To take advantage of support your config will be validated off course. In this first blogpost I'll go quickly over some architectural basics and in a second blog I'll elaborate more on the scaling/HA part of Nexenta.
This is how the basics of ZFS look like:

The Flash: DRAM storage. There's no dispute what so ever that this is the fastest storage of all. It's also the most expensive. Well, not always but I'll come back on that later. So what happens with RAM? Nexenta eats it ALL to put the data tables on (all but 1Gb for OS). This is good! In fact adding more RAM is the smartest thing to do in a Nexenta box.

SuperMan: SSD storage. This is used for 2 things. First is the level 2 caching for read (L2ARC). Obviously this holds the data that is most-used. And now a pretty sexy part: the ZIL. The ZIL keeps the writes coming from the applications a few seconds on the SSD so that it can be written sequentially to disk. This increases alot of extra bandwith we would have lost in a random write pattern.

The Juggernaut: power and volume. No more no less.

* note: the references to the comic figures are mine, not Nexenta's!

Commodity <> Certified hardware:

Commodity hardware means you can pick the hardware off the shelve. Is this really a smart idea? For a home/testing lab yes. For production? Maybe not. Therefore Nexenta has been working with multiple partners (ex. SuperMicro, DELL, ...) to create reference architectures. This is the brand new one for DELL:

Server: DELL PowerEdge R720 12G (around € 17,5k)

(2) Intel® Xeon® E5-2650 2.00GHz, 20M Cache, 8.0GT/s QPI, Turbo, 8C, 95W
(8) 16GB RDIMM, 1600 MHz, Standard Volt, Dual Rank, x4
Risers with up to 6, x8 PCIe Slots + 1, x16 PCIe Slot
Intel Ethernet X540 10Gb BT DP + I350 1Gb BT DP Network Daughter Card
Intel X520 DP 10Gb DA/SFP+ Server Adapter
iDRAC7 Enterprise with VFlash, 8GB SD Card
PERC H710 Integrated RAID Controller, 512MB NV Cache
(2) 146GB, SAS 6Gbps, 2.5-in, 15K RPM Hard Drive (Hot Plug)
(4) 200GB, SSD SAS 6Gbps, 2.5-in Hard Drive (Hot Plug)*
(2) PERC H810 RAID Adapter for External JBOD, 1Gb NV Cache

JBOD: PowerVault MD1200 (around €8.5k/node)

(2) PowerVault MD1200 Base
(2) 2 x 0.6M SAS Connector External Cable
(24) 600GB SAS 15k 3.5" HD

* : I took the SSDs from the online configuration tool but these should be STEC ZeusRAM disks.

Pricing:

If you count this up you have an Enterprise class tiered Scale-out NAS (10TB RAID6 or 7.5 TB RAID1+0) for less than €35k. Now imagine that the EqualLogic Hybrid model (PS6110XS 7xSSD, 17x10k) costs around 60k (10TB RAID 50) ... Important note: you're running now on the Community License which is free up to 18TB of storage (wth no support). I'll elaborate on the licensing in the second post. And I promised before that I would show you when RAM gets cheaper: enable dedupe! Nexenta puts the dedupe tables in the RAM (so it even eats more RAM) but this means you can do inline dedupe and so eliminate disks in the backend.

What's up next:

I'll show you how storage is presented to the hosts and how the different failover mechanisms work. EDIT: here it is finally!

In the mean time maybe you want to read this blog from a colleague of mine. Seems they are building freaky all-in-one boxes for home-lab, based on ZFS.

All-Flash solutions ready for commercial value?

2012-05-08T16:46:00.003+02:00

At Storage Field Day 1, the main theme was Solid State Storage. Last week I talked about Kaminario whom their best value proposition for me is the fact that they have an open architecture that can scale on any hardware platform. Another vendor I'd like to talk about is NimbusData. They were at the Solid State Storage Symposium and also in one of the dedicated vendor slots at Tech Field Day.

What is the value proposition of Nimbus Data? Making All-Flash solutions affordable without having to use deduplication. The last part seems a bit odd if you read it the first time but it has some value. There are 2 types of all-flash vendors out there: the ones that don't talk about $/TB (only $/IOPs) or the ones that tell you they are cheaper than disks because they have a huge deduplication ratio hence don't need volume (many SSDs). NimbusData pretends to be the 3rd kind that shows numbers.
Today NimbusData released a real life use-case at Mitsubishi Power Systems. The challenge here was replacing an existing disk-based solution. The applications (SAP/SQL) on top of the current array were stressing the entire system so much that it took 4 to 5 hours to do the nightly database loads and backup windows got missed. In the end the users tend to get frustrated on the overall SAP performance (how recognizable!).

With the S-Class solution, NimbusData was able to reduce the backup windows to complete within 20 minutes and the nightly database loads completed in less than 60-90 minutes. A very important sidenote: this is done on the same backend Storage Network. As the NimbusData solution is multi-protocol (SMB / NFS / iSCSI / FC / IB) they could easily integrate in the existing 4GB FC network, allowing the customer to upgrade the network design at a later point in time.

Throwing hardware against the problem is the easiest part. Paying for that hardware is another thing. NimbusData delivers all-flash at a price-point of $10k/TB (listprice). If I look at solutions I know we tend to see between $7.5k/TB for 15k disk solutions and $15k/TB for all-flash. I talked offline to the NimbusData sales people what happens in a real deal if a customer would go for an offer with a certified VAR (Value Added Reseller). At that point this even gets more interesting for the customer. You'll understand why I can not elaborate on that in public.

So how do they make it so cheap? They make their own flash drives with flash components from different suppliers! Below you see a picture of the 1TB raw EMLC disk (800GB netto) we got to see at Storage Field day 1. NimbusData claims that the SSD producers and vendors take way to much margin on their products keeping the prices artificially high. The fact that they are able to scale out with different disk/node sizes within the same storage pool is also something I like a lot (up to 100TB for S-Class, 500TB for E-Class).

In this video you see NimbusData CEO Tom Isakovich elaborates on other performance capabilities tested by Demartek.

My TakeAways:

As stated in the video I too believe that TIER1 disks (15k) will be replaced by SSDs pretty soon. The advantage of non-spinning hardware for high performance solutions is the way to go. But It will take us another year before more vendors drop prices on SSD.
We have a good, efficient and affordable solution for our high end applications. What about the lower tiers? I get the point NimbusData makes but we'll still need multiple storage solutions to manage in our environment.

Links:
NimbusData Corporate Website: LINK
NimbusData scaleout - S-Class: LINK
NimbusData scaleup - E-Class: LINK
Arjan Timmeman's (@ArjanTim) view on NimbusData: LINK

disclaimer: all transportation, catering and stay on TechFieldDay was taken care off by the sponsors you can find here through the organization of GestaltIT