–This has been revised based on some comments I’ve received since the original posting, check the comment thread if you’re interested what/why–

I came in this morning with an unusually clear diary, and took the liberty to check the newsfeeds for NetApp and EMC, this is when I came across an EMC press release entitled  ”EMC VNX SETS PERFORMANCE DENSITY RECORD WITH LUSTRE —SHOWCASES “NO COMPROMISE” HPC STORAGE“.

I’ve been doing some research on Lustre and HPC recently, and that claim surprised me more than a little, so I checked it out, maybe there’s a VNX sweetspot for HPC that I wasnt expecting. The one thing that stood out straight away was . “EMC® is announcing that the EMC® VNX7500 has set a performance density record with Lustre—delivering read performance density of 2GB/sec per rack” (highlight mine)

So today I found out that we’d broken a few records of our own few days ago, which was, at least from my perspective associated with surprisingly little fanfare with the associated press release coming out late last night. I’d like to say that the results speak for themselves, and to an extent they do. NetApp now holds the top two spots, and four out of the top five results on the ranking ladder. If this were the olympics most people would agree that this represents a position of PURE DOMINATION. High fives all round, and much chest beating and downing of well deserved delicious amber beverages.

So, apart from having the biggest number (which is nice), what did we prove ?

Benchmarks are interesting to me because they are the almost perfect intersection of my interests in both technical storage performance  and marketing and messaging. From a technical viewpoint, a benchmark can be really useful, but it only provides a relatively small number of proof points, and extrapolating beyond those, or making generalised conclusions is rarely a good idea.

For example, when NetApp released their SPC-1 benchmarks a few years ago, it proved a number of things

1. That under heavy load which involved a large number of random writes, a NetApp arrays performance remained steady over time

2. That this could be done while taking multiple snapshots, and more importantly while deleting and retiring them while under heavy load

3. That this could be done with RAID-6 and with a greater capacity efficiency as measured by  RAW vs USED than any other submission

4. That this could be done at better levels of performance than an equivalently configured  commonly used “traditional array” as exemplified by EMCs CX3-40

5. That the copy on write performance of the snapshots on an EMC array sucked under heavy load (and by implication similar copy on write snapshot implementations on other vendors arrays)

That’s a pretty good list of things to prove, especially in the face of considerable unfounded misinformation being put out at the time, and which, surprisingly is still bandied about despite the independently audited proof to the contrary. Having said that, this was not a “my number is the biggest”, exercise which generally proves nothing more than how much hardware you had available in your testing lab at the time.

A few months later we published another SPC-1 result which showed that we could pretty much doubl the numbers we’d achieved in the previous generation at a lower price per IOP with what was at the time a very competetive submission.

About two years after that we published yet another SPC-1 result with the direct replacement for the controller used in the previous test (3270 vs 3170). What this test didnt do was to show how much more load could be placed on the system, what it did do was to show that we could give our customers more IOPS at a lower latency with half the number of spindles .   This was the first time we’d submitted an SPC-1e result which foucussed on energy efficiency. It showed, quite dramatically how effective our FlashCache technology was under a heavy random write workload. Its interesting to compare that submission with the previous one for a number of reasons, but for the most part, this benchmark was about Flashcache effectiveness.

We did a number of other benchmarks including Spec-SFS benchmarks that also proved the remarkable effectiveness of the Flashcache technology, showing how it could make SATA drives perform as better than Fibre channel drives, or dramatically reduce the number of fibre channel drives required to service a given workload. There were a couple of other benchmarks done which I’ll grant were “hey look at how fast our shiny new boxes can run”, but for the most part, these were all done with configurations we’d reasonably expect a decent number our customers to actually buy (no all SSD configurations).

In the mean time EMC released some “Lab Queen” benchmarks, at first I thought that EMC were trying to prove just how fast their new X-blades were for processing CIFS and NFS traffic. They did this by configuring the back end storage system in such a rediculously overengineered way as to remove any possibility that they could cause a bottleneck in any way, either that or EMC’s block storage devices are way slower than most people would assume. From an engineering perspective I think they guys in Hopkington who created those X-blades did a truly excellent job, almost 125,000 IOPS per X-Blade using 6 CPU cores is genuinely impressive to me, even if all they were doing was  processing NFS/CIFS calls. You see, unlike the storage processors in a FAS or Isilon array, the X-Blade, much like the Network Processor in a SONAS system, or an Oceanspace N8500 relies on a back end block processing device to handle RAID , block checksums, write cache coherency and physical data movement to and from the disks, all of which is non-trivial work. What I find particularly interesting is that in all the benchmarks I looked at for these kinds of systems, the number of back end block storage systems was usually double that of the front end, which infers to me either that the load placed on back end systems by these benchmarks is higher than the load on the front end, or  more likely that the front end / back end architecture is very sensitive to any latency on the back end systems which means the back end systems get overengineered for benchmarks. My guess is after seeing the “All Flash DMX” configuration is that Celerra’s performance is very adversly affected by even slight increases in latency in the back end and that we start seeing some nasty manifestations of little law in these architectures under heavy load.

A little while later after being present at a couple of EMC presentations (one at Cisco Live, the other at a SNIA event, where EMC staff were fully aware of my presence), it became clear to me exactly why EMC did these “my number is bigger than yours” benchmarks. Ther marketing staff at corporate created a slide that compared all of the current SPC benchmarks in a way that was accurate, compelling and completely misleading all at the same time, at least as far as the VNX portion goes. Part of this goes back to the way that vendors, including I might say Netapp, use an availability group as a point of aggregation when reporting peformance numbers, this is reasonably fair as adding Active/Active or Active/Passive availability generally slows things down due to the two phase commit nature of write caching in modular storage environments. However, the configuration of the EMC VNX VG8 Gateway/EMC VNX5700 actually involves 5 separate availability groups (1xVG8 Gateway system with 4+1 redundancy, and and 4x VNX5700 with 1+1 redundancy). Presenting this as one aggregated peformance number without any valid point of aggregation smacks of downright dishonesty to me. If NetApp had done the same thing, then, using only 4 availabilty groups, we could have claimed over 760,000 IOPS by combining 4 of our existing 6240 configurations, but we didnt, because frankly doing that is in my opinion on the other side of the fine line where marketing finesse falls off the precipice into the shadowy realm of deceptive practice.

Which brings me back to my original question, what did we prove with our most recent submissions, well three things come to mind

1. That Netapp’s Ontap 8.1 Cluster mode solution is real, and it performs briliiantly

2. It scales linearly as you add nodes (more so than the leading competitors)

3. That scaling with 24 big nodes gives you better performance and better efficiency than scaling with hundreds of smaller nodes (at least for the SPEC benchmark)

This is a valid configuration using a single vserver as a point of aggregation across the cluster, and trust me, this is only the beginning.

As always, comments and criticism is welcome.

Regards

John

A couple of days ago I saw an email asking “what is the recommendation for maximum capacity utilization that will not cause performance degradation”. On the one hand this kind of question annoys me because for the most part it’s borne out of some the usual FUD which gets thrown at NetApp on a regular basis, but on the other, even though correctly engineering storage for consistent performance rarely, if ever, boils down to any single metric, understanding capacity utilisation and its impact on performance is an important aspect of storage design.

Firstly, for the record, I’d like to reiterate that the performance characteristics of every storage technology I’m aware of that is based on spinning disks decreases in proportion to the amount of capacity consumed.

With that out of the way, I have to say that as usual, the answer to the question of how does capacity utilisation affect performance is, “it depends”, but for the most part, when this question is asked, it’s usually asked about high performance write intensive applications like VDI, and some kinds of online transaction processing, and email systems.

If you’re looking at that kind of workload, then you can always check out good old TR-3647 which talks specifically about a write intensive high performance workloads where it says

The Data ONTAP data layout engine, WAFL®, optimizes writes to disk to improve system performance and disk bandwidth utilization. WAFL optimization uses a small amount of free or reserve space within the aggregate. For write-intensive, high-performance workloads we recommend leaving available approximately 10% of the usable space for this optimization process. This space not only ensures high-performance writes but also functions as a buffer against unexpected demands of free space for applications that burst writes to disk

I’ve seen other benchmarks using synthetic workloads where a knee in the performance curve begins to be seen at between 98% and 100% of the usable capacity after WAFL reserve is taken away, I’ve also seen performance issues when people completely fill all the available space and then hit it with lots of small random overwrites (especially misaligned small random overwrites). This is not unique to WAFL, which is why it’s a bad idea generally to fill up all the space in any data structure which is subjected to heavy random write workloads.

Having said that for the vast majority of workloads you’ll get more IOPS per spindle out of a netapp array at all capacity points than you will out of any similarly priced/configured box from another vendor

Leaving the FUD aside, (the complete rebuttal of which requires a fairly deep understanding of ONTAP’s performance achitecture)  when considering capacity and its effect on performance on a NetApp FAS array it’s worth keeping the following points in mind.

1. For any given workload, and array type you’re only ever going to get a fairly limited number transactions per 15K RPM disk, usually less than 250
2. Array performance is usually determined  by how many disks you can throw at the workload
3. Most array vendors bring more spindles to the workload by using RAID-10 which uses twice the amount of disks for the same capacity, NetApp uses RAID-DP which does not automatically double the spindle density
4. In most benchmarks (check out SPC-1), NetApp uses all but 10% of the available space (in line with TR-3647) which allows the user to use approximately 60% of the RAW capacity  while still achieving the same kinds of IOPS/drive that more other vendors are only able to do using 30% of the RAW capacity. i.e at the same performance per drive we offer 10% more usable capacity than the other vendors could theoretically attain using RAID-10.

The bottom line is, that even without dedupe or thin provisioning or anything else you can store twice as much information in a FAS array for the same level of performance as most competing solutions using RAID-10

While that is true, it’s worth mentioning it does have one drawback. While the IOPS/Spindle is more or less the same, the IOPS density measured in IOPS/GB on the NetApp SPC-1 results is about half that of the competing solutions, (same IOPS , 2x as much data = half the density). While that is actually harder to do because you have a lower cache:data ratio, if you have an application that requires very dense IOPS/GB (like some VDI deployments for example), then you might not be allocate all of that extra capacity to that workload.  This in my view gives you three choices.

1. Don’t use the extra capacity, just leave it as unused freespace in the aggregate which will make it easier to optimise writes
2. Use that extra capacity for lower tier workloads such as storing snapshots or a mirror destination, or archives etc, and set those workloads to a low priority using FlexShare
3. Put in a FlashCache card which will double the effective number of IOPS (depending on workload of course) per spindle, which is less expensive and resource consuming than doubling the number of disks

If you dont do this, then you may run into a situation I’ve heard of  in a few cases where our storage efficiencies allowed the user to put too many hot workloads on not enough spindles, and unfortunately this is probably the basis for the  “Anecdotal Evidence”  that allows the Netapp Capacity / Performance FUD to be perpetuated. This is innacurate because it has less to do with the intricacies of ONTAP and WAFL, and far more to do with systems that were originally sized for a workload of X having a workload of 3X placed on them because there was still capacity available on Tier-1 disk capacity, long after all the performance had been squeezed out of the spindles by other workloads.

Keeping your storage users happy, means not only managing the available capacity, but also managing the available performance. More often than not, you will run out of one before you run out of the other and running an efficient IT infrastructure means balancing workloads between these two resources. Firstly this means you have to spend at least some time measuring, and monitor both the capacity and performance of your environment. Furthermore you should also set your system up to it’s easy to migrate and rebalance workloads across other resource pools, or be able to easily add performance to your existing workloads non disruptively which can be done via technologies such as Storage DRS in vSphere 5, or ONTAP’s Data motion and Virtual storage tiering features.

When it comes to measuring your environment so you can take action before the problems arise, NetApp has a number of excellent tools to monitor the performance of your storage environment. Performance Advisor gives you visualization and customised alerts and thresholds for the detailed inbuilt performance metrics available on every FAS Array, and OnCommand Insight Balance provides deeper reporting and predictive analysis of your entire virtualised infrastructure including non-NetApp hardware.

Whether you use NetApp’s tools or someone elses, the important thing is that you use them, and take a little time out of you day to find out which metrics are important and what you should do when thresholds or high watermarks are breached. If you’re not sure about this for your NetApp environment, feel free to ask me here, or better still open up a question in the Netapp communities which has a broader constituency than this blog.

While I appreciate that it’s tempting to just fall back to old practices, and overengineer Tier-1 storage so that there is little or no possibility of running out of IOPS before you run out of capacity, this is almost always incredibly wasteful and has in my experience resulted in storage utilisation rates of less than 20%, and  drives the costs/GB for “Tier-1″ storage to unsustainable and uneconomically justifiable heights. The time may come when storage administrators are given the luxury of doing this again, or you may be in one of those rare industries where cost is no object, but unless you’ve got that luxury, it’s time to brush up on your monitoring and storage workload rebalancing and optimisation skills. Doing more with less is what it’s all about.

As always, comments and contrary views are welcomed.

Preamble

During a discussion I had at the SNIA blogfest, I mentioned that I’d written a whitepaper around archiving and I promised that I’d send it on. It took me a while to get around to this, but I finally dug it out from my archive, which is implemented in a similar way to the example policy at the bottom of this post.,  it only took me about a minute to search and retreive it once I’d started the process of looking for it from a FAS array that was far enough away from me to incur a 60ms RTT latency. Overall I was really happy with the result.

The document was in my archive because I wrote it almost two years ago, since then a number of things have changed, however the fundamental principals have not, I’ll work on updating this when things less busy, probably sometime around January ’11. On a final note, because I wrote this a couple of years ago when my job role was different than it is today, this document is considerably more “salesy” than my usual blog posts, it shouldnt be construed as a change in direction for this blog in general.

Introduction

There are a number of approaches that can be broadly classified as some form of Archiving, including Hierarchical Storage Management (HSM), and Information Lifecycle Management (ILM).  All of these approaches aim to improve the IT environment by

• Lowering Overall Storage Costs
• Reducing the backup load and improving restore times
• Improving  Application Performance
• Making it easier to find and classify data

The following kinds of claims are common in the marketing material promoted by vendors of Archiving software and hardware

“By using Single Instance Storage, data compression and an ATA-based archival storage system (as opposed to a high performance, Fibre Channel device), the customer was able to reduce storage costs by $160,000 per terabyte of messages during the first three years that the joint EMC / Symantec solution was deployed. These cost savings were just the beginning, as the customers were also able to maintain their current management headcount despite a 20% data growth and the time it took to restore messages was drastically reduced. By archiving the messages and files, the customer was also able to improve electronic discovery retrieval times as all content is searchable by keywords.”

These kinds of results while impressive assume a number of things that are not true in NetApp implementations

• A price difference between primary storage and archive storage of over $US160,000 per TB.
• Backup and restores are performed from tape using bulk data movement methods
• Modest increases in storage capacities require additional headcount

In many NetApp environments, the price difference between the most expensive tier of storage and the least expensive simply does not justify the expense and complexity of implementing an archiving system based on a cost per TB alone.

For file serving environments, many file shares can be stored effectively on what would be traditionally thought of as “Tier-3” storage with high-density SATA drives, RAID-6 and compression / deduplication. This is because unique NetApp technologies such as WAFL and RAID-DP provide the performance and reliability required for many file serving environments. In addition, the use of NetApp SnapVault replication based data protection, for backup and long term retention means that full backups are no longer necessary. The presence or absence of the kinds static data typically moved into archives has little or no impact on the time it takes to perform backups, or make data available in the case of disaster.

Finally, the price per GB and IOPS for NetApp storage has fallen consistently in line with the trend in the industry as a whole. Customers can lower their storage costs by purchasing and implementing storage only as required. NetApp FAS array’s ability to non-disruptively add new storage, or move excess storage capacity and I/O from one volume to the other within an aggregate makes this approach both easy, and practical.

While the benefits of archiving for NetApp based file serving environments may be marginal, archiving still has significant advantages for email environments, particularly Microsoft Exchange. The reasons for this are as follows

1. Email is cache “unfriendly” and generally needs many dedicated disk spindles for adequate performance.
2. Email messages are not modified after they have been sent/received
3. There is a considerable amount of “noise” in email traffic (spam, jokes, social banter etc)
4. Small Email Stores are easier to cache, which can significantly improve performance and reduce the hardware requirements for both the email servers and the underlying storage
5. Email is more likely to be requested during legal discovery
6. Enterprises now consider Email to be a mission critical application and some companies still mandate a tape backup of their email environments for compliance purposes.

Choosing the right Archive Storage

It’s about the application

EMC and NetApp take very different approaches to archive storage, each of which works well in a large number of environments. An excellent discussion on the details of this can be found in the NetApp whitepaper WP-7055-1008 Architectural Considerations for Archive and Compliance Solutions. For most people however, the entire process of archive is driven not at the storage layer, but by the archive applications. These applications do an excellent job of making the underlying functionality of the storage system transparent to the end user, however the user is still exposed to the performance and reliability of the storage underlying the archives.

Speed makes a difference

Centera was designed to be “Faster than Optical” and while it has surpassed this relatively low bar, its performance doesn’t come close to even the slowest NetApp array. This is important, because the amount of data that can be pushed onto the archive layer is determined not just by IT policy, but also by user acceptance and satisfaction with the overall solution. The greater the user acceptance, the more aggressive the archiving can be, which results in lower TCO and faster ROI.

Protecting the Archive

While the archive storage layer needs to be reliable, it should be noted that without the archive application and its associated indexes, the data is completely inaccessible, and may as well be lost. It might be possible to rebuild the indexes and application data from the information in the archive alone, often this process may be unacceptably long. Protecting the archive involves protecting the archive data store, the full text indexes, and the associated databases in a consistent manner at a single point in time.

Migrating from an Existing Solution

Many companies already have archiving solutions in place, but would like to change their underlying storage system to something faster and more reliable. Fortunately archiving applications build the capability to migrate date from one kind back-end storage to another into their software. The following diagrams show how this can be achieved for EmailXtender and DiskXtender to move data from Centera to NetApp.

Some organizations would prefer to completely replace their existing archiving solutions including hardware and software. For these customers NetApp collaborates with organizations such as Procedo (www.procedo.com), to make this process fast and painless.

SnapVault

As mentioned previously, the cost and complexity of traditional archiving infrastructure may not add sufficient value to a NetApp file-serving environment, as many of the problems it solves are already addressed by core NetApp features. This does not mean that some form of storage tiering could not or should not be implemented on FAS to reduce the amount of NetApp primary capacity.

One easy way of doing this is by taking advantage of the flexibility of the built in backup technology. This is an extension of the “archiving” policy used by many customers, where the backup system is used for archive as well.  The approach of mixing backup and archive is rightly discouraged by most storage management professionals, the reasons for doing so in traditional tape based backup environments don’t apply.

The reasons for this are

• Snapshot and replication based backups are not affected by capacity as only changed blocks are ever moved or stored
• The backups are immediately available, and can be used for multiple purposes
• Backups are stored on high reliability disk in space efficient manner using both non-duplication and de-duplication techniques
• Files can be easily found via existing user interfaces such as Windows Explorer or external search engines

In general, SnapVault destinations use the highest density SATA drives with the most aggressive space savings policies applied to them. These policies and techniques, which may not be suitable for high performance file sharing environments, provide the lowest cost per TB of any NetApp offering. This combined with the ability to place the SnapVault destination in a remote datacenter may relieve the power, space and cooling requirements of increasingly crowded datacenters.

An example policy

Many companies file archiving requirements are straightforward, and do not justify the detailed capabilities provided by archiving applications. For example, a company might implement the following backup and archive policy

• All files are backed up on a daily basis with daily recovery points kept for 14 days, weekly recovery points will be kept for two months and monthly recovery points kept for seven years.
• Any file that has not been accessed in the last sixty days will be removed from primary storage and will need to be accessed from the archive

This is easily addressed in a SnapVault environment through the use of the following

• Daily backups are transferred from the primary system to the SnapVault repository
• Daily recovery points (snapshots) are kept on both the primary storage system and the SnapVault repository for 14 days
• Weekly recovery points (snapshots) are kept only on the SnapVault repository
• Monthly recovery points (snapshots) are kept only on the SnapVault repository
• A simple shell script/batch file is executed after each successful daily backup which deletes any file from the primary volume that has not been accessed in thirty days
• Users are allocated a drive mapping to their replicated directories on the SnapVault destination.
• Optionally the Primary systems and SnapVault repository may be indexed by an application such as the Kazeon IS1200, or Google enterprise search.

Users then need to be informed that old files will be deleted after thirty days, and that they can access backups of their data, including the files that have been deleted from primary storage by looking through the drive that is mapped to the SnapVault repository, or optionally via the enterprise search engines user access tools.

By removing the files from primary storage, instead of the traditional “stub” approach favoured by many archive vendors, the overall performance of the system will be improved by reducing the metadata load, and users will be able to more easily find active files by having fewer files and directories on the primary systems.

Conclusion

Many Organisations archiving requirements can be met by simply adding additional SATA disk to the current production system replicated via SnapMirror to the current DR system – rather than managing separate archive platforms.

This architecture provides flexibility and scalability over time and reduces management overhead. Tape can also be used for additional backup and longer term storage if required. SnapLock provides the non-modifiable WORM like capability required of an archive without additional hardware (a software licensable feature, see more detail at http://www.netapp.com/us/products/protection-software/snaplock.html ).

The last few months have been interesting for me,  as my new job role involves a lot of work with alliance partners, many of whom either didn’t know anything about NetApp, or where they did know something it was along the lines of “Oh yeah, the NAS company”. In many respects, it’s a lot easier to explain what we do when someone has an open mind, as pre-conceived notions are often hard to budge, and telling someone they’re wrong is rarely a good way to start a trusted relationship. Even though I report up through our local director of marketing, my soul is still that of an engineer, so when it comes to describing what NetApp does, and why that’s important I tend to go straight to “Well, we still sell NAS, and that’s a big part of what we do, but we really sell is Unified Storage” at which point I expect to see the “and I should care about this because …” look

I’ve been seeing this look quite a bit recently, mostly because many of the people I speak to also get briefs from other storage vendors, and they too have suddenly started talking about “Unified Storage” without really understanding it or explaining its relevance to datacenter transformation. A good case in point was the opportunity I had to speak at the local VMware seminar series where I shared a stage with VMware, Cisco, and EMC. All of us got our 7.5 minutes  to explain how we helped accelerate our customers journey to the cloud. VMware went first, followed by EMC, then Cisco and then me..

I’d prepared two slides for my 7 minutes focussing on our key differentiators, Unified storage, tight VMware integration with advanced storage features, deduplication and storage efficiency, Secure Multi-Tenancy, Cisco validated designs, Backup and recovery, and waited happily to see if EMC would come out with their usual pitch.

Boy, was I surprised … EMC’s pitch was Unified storage, deduplication, tight VMware integration with advanced storage features, deduplication and  storage efficiency, security and Cisco validated designs … what the ????, had I suddenly slipped into a parallel universe ? Had EMC, a company fairly well known for pushing seven different kinds of storage with forklift upgrades  suddenly capitulated and acknowledged that the approach NetApp had been pushing for so many years was actually right ? Was Chuck Hollis about to come on stage and apologise for blatant manipulation of social media and comment filtering ?

Now while I could have picked holes in their story by pointing out that at least from a VMware perspective they don’t have deduplication, that their advanced integration with VAAI hadn’t been released, there was no Cisco Validated Design for vBlock, and that the RSA stuff had no integration at the storage layer, nobody is really interested in hearing vendors denigrate each other, and I only had 7 minutes to figure out how to show our unique ability to help customers in the face of the most shameless “me too” campaign I’ve ever seen.  During that 7 minutes there was one thing that really struck me. EMC has no real concept of why unified storage is important. Their concept of unified storage was something that allowed connection by Fibre Channel, iSCSI, CIFS and NFS and had a nice GUI. Having worked at NetApp for a number of years, I was surprised at how they’d missed the point completely. Almost everyone at NetApp knows that these are good features to have (we’ve had them for over 10 years now), but we also know that by themselves, they have only limited benefits. I’ve had a little while now to think about this, and it’s become clear to me that for other vendors, Unified storage is not a strategic direction, but a tactical response to NetApp’s continued success in gaining market share. This becomes even more obvious by taking a look at their storage portfolios

Now, if you match one of those arrays against the workload they were designed for, you’ll probably get a pretty good result. In a static, reasonably predictable environment without much change, you could make a reasonable argument that this was the best approach to take. You built a silo around an application or function, and purchased the equipment that matched that function. I’ve seen more than one customer that had every product in a vendors portfolio, and seem to be fairly happy, or at least have been until fairly recently.

The problem with these narrow silo’ed approaches is that each silo creates new inefficiencies and dedicated areas of whitespace in both capacity, performance. For example, there is no way of taking excess capacity allocated to a  backup to disk appliance and start using it for CIFS home directories, nor is there a way of taking the excess IOPS capability of temporarily idle disk archive and allocate those IOPS to another application undergoing an unusual workload spike such as a VDI bootstorm.

But for me, the biggest area of waste is that of management. Each of these silo’s tends to get its own set of administrators and workflows, each of which may, or may not work in harmony with the other. Most of us have experienced the bitterness and waste of IT turf wars, and the traditional vendors not only encourage, but depend on and help  maintainin these functionality silos, as it allows a divide and conquer sales model that benefits the vendor far more than the customer. If there was a book entitled “How to build an inflexible and wasteful IT infrastructure”, I imagine that encouraging and spreading “IT functionality silos” would fill up the first few chapters. Even though there are a bunch of people who have been quite happy with this status quo, and the business processes from budgeting and product selection all the way through to procurement and training that entrenches this model, things are changing, and they’re changing a lot faster than I thought they would.

A lot of credit for this change has to go to vendors like Microsoft, Cisco and VMware whose products have blurred the lines of these traditional silos. Virtualisation at both the compute and network layers have driven the kinds of cross functional change CIO’s have been crying out for, and in the wake of this, Unified storage finds its natural fit ; not because of its support multiple protocols, but because these environments require the kind of  workload agility and managment simplicity that only a truly Unified storage offering can fully satisfy.

But it’s not just server and desktop virtualisation and other forms of shared infrastructure where unified storage is a natural fit. Almost any “multi-part” or landscape style application can benefit too, not just because of the flexibility and efficiency, but more importantly because of the fact that these environments are really hard to protect effectively. A really good example of this is Enterprise Content Management Applications such as FileNet and SharePoint

Typically these applications have

• Content Servers
• Business Process Workflow Engines / Servers
• Database servers
• Index Servers
• Content Servers

In a large installation, there will be many of these servers and multiple databases, indexes and content repositories to cater for scalability and in some cases, the tyranny of distance (latency is forever).

In a “traditional” silo’d model this data would be stored on two or possibly three different kinds of storage arrays, each with its own method of backup and replication, most of which depend on some form of “bulk copy” backup method as the primary form of logical data protection. The effect of this is that backing up these ECM systems on traditional storage architectures is almost impossible. While I’ve been talking to customers about this for a few years, recently there seems to have been a big increase in customers seeing these problems. In one case a design review for a Petabyte scale SharePoint implementation identified that if a critical index was lost the entire infrastructure would be effectively unrecoverable, and that there was no effective backup capability. In another discussion I had today around redesigning data protection, a brief mention on ECM created more interest than almost anything else simply because of the difficulty of backing their Documentum system.

Truly unified storage not only allows data to be stored using multiple protocols, and provides rich functionality like deduplication and compliant WORM storage which makes it a logical choice for  ECM solutions, but more importantly it also provides a single integrated method for protecting that data in a way that is application-consistent without the need for a “cold” backup. And, you guessed it, NetApp can do that with ease, whereas other vendors’ versions of what they are calling unified storage would find that challenging (to be kind).

In the next few posts, I’ll take a deeper dive into exactly what NetApp does for Enterprise Content Management, with a focus on why Unified Storage is such a good match, and what we do to protect a company’s most important data assets.

I’d intended writing about Megacaches in both the previous post, as well as this one, but interesting things keep popping up that need to be dealt with first. This time it’s an article at information week With VDI, Local Disk Is A Thing Of The Past. In it Elias Khnaser outlines the same argument that I was going to make after I’d dealt with the technical details of how NetApp optimises VDI deployments.

I still plan to expand on this with posts on Megacaches, single instancing technologies, and broker integration, but Elias’ post was so well done that I thought it deserved some immediate attention.

If you havent done so already, check out the article before you read more here, because the only point I want to make in this uncharacteristically small post is the following

The capital expenditure for storage in a VDI deployment based on NetApp is lower than one based on direct attached storage.

This is based on the following

Solution 1: VDI Using Local Storage – Cost

$614,400

Solution 2 : VDI Using HDS Midrange SAN – Cost

$860,800, with array costs of approx $400,000

Solution 3 : VDI Using FAS 2040 – Cost

860,000 – 400,000 + (2000 * $50) = $560,000

You save $54,000 (about 10% overall) compared to DAS and still get the benefits of shared storage. That’s $56,000 you can spend on more advanced broker software or possibly a trip to the Bahamas.

Now if you’re wondering where I got my figures from, I did the same sizing exercise I did in Part 7 of this post but using 12 IOPS per user and using 33:63 R:W ratio. I then came up with a configuration and asked one of my colleagues for a street price. The figure came out to around $US50/desktop user for an NFS deployment, which is inline with what NetApp has been saying for about our costs for VDI deployments for some time now.

Even factoring in things like professional services, additional network infrastructure, training etc, you’d still be better off from a up-front expenditure point of view using NetApp than you would with internal disks.

Given the additional OpEx benefits, I wonder why anyone would even consider using DAS, or even for that matter another vendors SAN.