Author Archives: chrismaki

ADP(v1) and ADPv2 in a nutshell, it’s delicious!

Ever since clustered Data ONTAP went mainstream over 7-Mode, the dedicated root aggregate tax has been a bone of contention for many, especially for those entry-level systems with internal drives. Can you imagine buying a brand new FAS2220 or FAS2520 and being told that not only are you going to lose two drives as spares, but also another six to your root aggregates? This effectively left you with four drives for your data aggregate, two of which would be devoted to parity. I don’t think so. Now, this is a bit of an extreme example that was seldom deployed. Hopefully you had a deployment engineer who cared about the end result and would use RAID-4 for the root aggregates and maybe not even assign a spare to one controller, giving you seven whole disks for your active-passive deployment. Still, this was kind of a shaft. In a 24-disk system deployed active-active, you’d likely get something like this:

Traditional cDOT

Enter ADP.

In the first version of ADP introduced in version 8.3, clustered Data ONTAP gained the ability to partition drives on systems with internal drives as well as the first two shelves of drives on All Flash FAS systems. What this meant was the dedicated root aggregate tax got a little less painful. In this first version of ADP, clustered Data ONTAP carved each disk into two partitions: a small one for the root aggregates and a larger one for the data aggregate(s). This was referred to as root-data or R-D partitioning. The smaller partition’s size depended on how many drives existed. You could technically buy a system with fewer than 12 drives, but the ADP R-D minimum was eight drives. By default, both partitions on a disk were owned by the same controller, splitting overall disk ownership in half.

8.3 ADP, R-D


You could change this with some advanced command-line trickery to still build active-passive systems and gain two more drive partitions’ worth of data. Since you were likely only building one large aggregate on your system, you could also accomplish this in System Setup if you told it to create one large pool. This satisfied the masses for a while, but then those crafty engineers over at NetApp came up with something better.

Enter ADPv2.

Starting with ONTAP 9, not only did ONTAP get a name change (7-Mode hasn’t been an option since version 8.2.3), but it also gained ADPv2 which carves the aforementioned data partition in half, or R-D2 (Root-Data,Data) sharing for SSDs. Take note of the aforementioned SSDs there, as spinning disks aren’t eligible for this secondary partitioning. In this new version, you get one drive back that you would have allocated to be a spare, and you also get two of the parity drives back, lessening the pain of the RAID tax. With a minimum requirement of eight drives and a maximum of 48, here are the three main scenarios for this type of partitioning.

12 Drives:

ADPv2, R-D2 ½ shelf

24 Drives:

ADPv2, R-D2 1 shelf

48 Drives:

ADPv2, R-D2 2 shelves

As you can see, this is a far more efficient way of allocating your storage that yields up to ~17% more usable space on your precious SSDs.

So that’s ADP and ADPv2 in a nutshell—a change for the better. Interestingly enough, the ability to partition disks has lead to a radical change in the FlashPool world called “Storage Pools,” but that’s a topic for another day.

NetApp refreshes entire line of FAS and AFF platforms

Today NetApp announced a complete revamping of both the FAS and AFF lines and with it a divergence in model numbers. My favourite improvement is that NetApp has changed the way FlashCache gets delivered; now all FAS platforms can take advantage of FlashCache using an M.2 NVMe device onboard the controller, even the entry-level models; in fact, it’s standard on all models. In the realm of connectivity, both the top-end FAS as well as all AFFs can now offer not only 40GbE but 32Gb FC as well, first to market for both of these.

Without further ado, here are the new models in the FAS line:

  • FAS2620 and FAS2650
    • Appears to be the same 2RU enclosure as the FAS2240-2, FAS2552, and DS2246, likely with an upgraded mid-plane.
    • FAS2620 holds 12 large form factor (3.5″ NL-SAS/SSD) drives internally
    • FAS2650 holds 24 small form factor (2.5″ SAS/SSD) drives internally
    • Both models come with 1TB of FlashCache
  • FAS8200
    • Appears to be the same 3RU enclosure as the FAS8020
    • 1TB of FlashCache is now standard, upgradeable to 4TB
  • FAS9000
    • This all-new chassis separates the I/O from the controller so there are no more onboard ports and all I/O is done using PCIe cards, 10 slots per node.
    • 2TB of FlashCache are now part of the standard configuration, upgradeable to 16TB.

And the new AFF line now consists of:

  • A300 (Same chassis as FAS8200)
  • A700 (Same chassis as FAS9000)

Strictly the numbers*:

Model RU RAM NVRAM (NVMEM) Max HDD (SDD) Max Flash Cache Max Flash Pool Onboard UTA2 Onboard 10GbE Onboard 10GbE Base-T Onboard 12GB SAS PCIe Expansion Slots Cores
FAS2620 2 64GB (8GB) 144 1TB 24TB 8 4 4 N/A N/A 12
FAS2650 2 64GB (8GB) 144 1TB 24TB 8 4 4 N/A N/A 12
FAS8200 3 256GB 16GB 480 4TB 48TB 8 4 4 4 4 32
FAS9000 8 1024GB 64GB 1440 (480) 16TB 144TB N/A N/A N/A N/A 20 72
A300 3 256GB** 16GB 384 N/A N/A 8 4 4 4 4 32
A700 8 1024GB 64GB 480 N/A N/A N/A N/A N/A N/A 20 72
  • *Numbers are per HA pair
  • **16GB Carved out for NVLOGS

Performance Improvements

The FAS2600 comes with 3 times as many cores, twice as much memory and >3 times the NVMEM than that of the FAS2500 and brings 12Gb SAS and 1TB of NVMe FlashCache is expected to perform 200% faster than its predecessor running 8.3.x, making the entry-level line of controllers smoking fast. The 8200 has twice as many cores and four times as much memory as the FAS8040 and also comes with 12Gb SAS and 1TB of FlashCache, making it roughly 50% faster.

The new top-end model, the FAS9000 goes modular, decoupling I/O from the controllers. This performance monster which has 2TB of FlashCache standard and 20 PCIe slots for I/O is expected to run 50% faster than the FAS8080 on 8.3.x. A cluster of 24 FAS9000 nodes (12 HA pairs) scales up to as much as 172PB.

FAS9000 AFF A700 Chassis

Here’s how the new models map to the old:

New FAS platforms

As for the new AFF models, the A300 should get about 50% more throughput than AFF8040 running 8.3.1 while the A700 aims to replace the dual chassis AFF8080, saving four precious rack units but still providing 100% more IOPS, in fact it should be able to handle about double the workload at half the latency.

Oracle testing

And here’s how the new AFFs line up with the existing ones:

AFF model alignmentThe new lineup, both FAS and AFF are definitely addressing some concerns; FlashCache not only available throughout the FAS line but standard as well is a move in the right direction as is the addition of 12Gb SAS. The introduction of both 40GbE and 32Gb FC into the mid-range and upper models of both lines should provide the fire hose required to deliver all that new controller and storage back-end performance. The two new AFF model numbers lead me to believe that they may be leaving room in the middle to add models to the line.

While ADP has been around for a while and is a great work around to dedicated root aggregates, I would love to see NetApp move away from root aggregates completely and do something with M.2. I’ll keep my fingers crossed for this one, but won’t hold my breath either.

ONTAP 9.0 is here.

That’s right folks, not the 8.4 you were thinking was next, but straight to 9. With the jump to 9 in the versioning also comes a bit of rebranding. When referring to the OS of your FAS, you can finally simply say ONTAP, no more qualifying that with “7-mode” or “clustered” or even prefixing it with “Data”. Alongside this new version comes two variants, one that runs in the cloud and one that you can run in your VMware environment. ONTAP Cloud and ONTAP Select respectively. Currently ONTAP Cloud is AWS only, but all signs point to an Azure release in the very near future. ONTAP Select picks up where Edge left off.

I’ve had some helm time with the new version and the first thing you’ll notice is that System Manager has been cleaned up and rearranged. When you first login, you’ll now get the following dashboard with a quick view of your cluster:

Screen Shot 2016-05-30 at 4.58.02 PM

This dashboard is good for a quick glance at performance and capacity, but clicking around the other tabs still leaves something to be desired on the usability front, but only because there is just so much available in this interface. I feel like I’ve got the “advanced view” option permantly checked off. Personally I don’t mind all the various tabs and sub-tabs, but for your day-to-day operator, most of the options available aren’t necessary.

Moving over to the technical side of the equation, ONTAP 9 brings with it a few new exciting features. First of all, support for the new 15.3TB SSDs makes its way into the payload, as well as RAID-TEC triple-parity protection. As far as I know, NetApp will be first to market with these 15.3’s and I can’t wait to see them in the field. RAID-TEC, or RAID Triple Erasure Coding will really help out with the larger disk sizes. While it won’t be mandatory for the large SSDs, I highly recommend it for spinning drives >= 6TB. These drives currently have a max RAID group size of 14, but the introduction of RAID-TEC will increase that to 28 drives. This will not only double your RAID protection level and decrease rebuild times,  but most importantly the RAID tax won’t be so bad in the larger deployments. If you’ve already got these large drives deployed, you can move to RAID-TEC and larger RAID groups provided you have the disks to add to the aggregate.

In the realm of performance, NetApp is claiming a 60% increase in IOPS over 8.3.1, as well as the introduction of “Headroom for visibility of performance capacity”. What this means is that at a glance, you should be able to see how much more performance is left in your cluster. NetApp has also introduced a new data-reduction technology called Data Compaction. With this latest addition to the existing data reduction tricks, namely deduplication, compression, cloning and thin-provisioning NetApp is now boasting a 4:1 data reduction number and is backing it with a guarantee.

Finally, two more feature introductions for the compliance-minded folks out there. First you’ll be happy to hear SnapLock® software is back, and for those not looking to introduce the cost and complexity of an external key manager for NSE drives, NetApp has introduced an onboard key manager.


Make sure to check out some of the other posts on this subject by my fellow NetApp A-Team members:

***Note: ONTAP 9 is not out yet, but the details are. The exact release date is not public yet.

ONTAP 8.3.2 is GA!

Just a quick note now that 8.3.2 is now GA and here’s a few features that I may have missed in my combined post on 8.3.1 and 8.3.2 or the details of which just weren’t available to me. I’ll start by reiterating some of the highlights in point form:

  • Copy Free Transition (CFT)
  • In-line DeDupe (on AFF)
  • In-place, adaptive compression

Those were some of the more poignant features, but here’s some others that I either didn’t know about or just weren’t to me:

  • MetroCluster distance increased by 50% bringing it to 300KM
  • Oracle NFS workload provisioning for All Flash FAS (AFF)
    • Using a Quick Start guide, and a wizard in the on-board OnCommand System Manager, the claim is you can have your new AFF cabled and serving Oracle over NFS in under 15 minutes.

The real beauty in this release however is in regards to CFT. I’ve said it before, but I’m still impressed by this feature. Basically you can stand up a new cDOT system with it’s own root aggregates and connect your 7-mode disk to it (yes there’s caveats) and with some 7MTT magic smoke, your data is now being served out of your shiny, new cluster-mode environment. Previously 7MTT (7-mode Transition Tool) only supported source data in the 8.1.x code line, but with this new release, 7MTT now supports 7-Mode systems running Data ONTAP 8.1.4P4 – 8.1.4P9 and Data ONTAP 8.2.1 or later. Also, in 8.3.2RC, CFT would only work on a net-new system with only its root aggregates, but in the 8.3.2GA, CFT now supports importing your 7-mode disk on to 8.3.2GA systems with pre-existing data aggregates and volumes. Here are all the permutations that you can now leverage CFT to import your 7-mode data:

  • Import 7-Mode disk shelves in the following ways:
    • Import disk shelves from a 7-Mode HA pair to a new HA pair in a new cluster.
    • Import disk shelves from a 7-Mode HA pair to a new HA pair in an existing cluster that has additional data-serving nodes.
    • Import disk shelves from a 7-Mode HA pair to an HA pair that has data aggregates in an existing cluster that is serving data.
    • Import disk shelves from an HA pair that contains volumes in a volume SnapMirror relationship to an HA pair in a new or existing cluster.
      You must manually create the cluster peer relationship after transition; however, a rebaseline transfer is not required, and you can retain the SnapMirror relationship after transition.

*UPDATE: The above is actually more a function of 7MTT 2.3, but you need it for CFT anyway.

For all the gory details of CFT and how awesome it is, go here for your copy of the Copy-Free Transition Guide.

What does all this actually mean? It means that soon we can finally stop referring to “it” as either 7-mode or cluster-mode and just refer to “it” as ONTAP or Data ONTAP again.

Have Your Pi and Eat It Too!

Last year I gained access to an arcade cabinet that had been outfitted with a standard consumer PC running Windows and being the front end to a handful of emulators. The cabinet itself is actually a custom build, not a repurposed one from years ago. This was purchased from Arcade Time Machine and it is pretty awesome. That however is not the point of this post, what I am doing is tracing back the root of my RaspberryPi obsession. Talking about the cabinet with some co-workers, they were telling me that you can do all of that on a RaspberryPi, and there was actually a community around exactly that. The “solution” is called RetroPie and you can read more about that over there. I quickly visited Amazon and put together a list of components  which included the following:

  • Raspberry Pi 2 Model B
  • Power Supply and Case
  • SD Card
  • WiFi dongle
  • HDMI Cable
  • 2 × Logitech F710 gaming controllers
  • Logitech K400 wireless keyboard

Pretty much everything I needed to get my own RetroPie gaming system up and running at home.

Talking with various co-workers around the country over our internal Slack, I found that apparently one guy had built his RetroPie and everyone else had pretty much just gotten an image of his SD card. I decided that I would also go down this route. I got the image burnt and after a bit of tweaking I had a mostly functioning system. I played some of the old 8-bit games that I used to love, Pyros, Shinobi and Mario Brothers to name a few. Some things just didn’t work how I wanted them to however, so I figured I’d have to build my own RetroPie image. While I haven’t tried installing the RetroPie packages on top of Raspbian myself yet, that is the next step. Currently my Pi boots into RetroPie using the image you can download from RetroPie directly but I just can’t get it to work and I think it’s because I don’t know enough about what’s going on behind the GUIs and the only way to figure it out is to install everything myself and not rely on an image file.

Between the co-worker-supplied image and now however, I’ve done a few other things with my Pi. One of the neat things you can do with it is turn it into a digital media set-top box that can replace or supplement your AppleTV, Chromecast or Roku. I did this by installing Kodi (formerly XBMC) on it. Now I was already using XBMC on my jailbroken ATV2, so I already had some expectations which it pretty much met. The Pi however appears to run Kodi much faster than the ATV2, possibly due to it being newer, faster hardware. I don’t really know or care actually as it was only a test and I still rely on my ATV2 for day-to-day viewing, at least until I buy a supplementary Pi to devote to this. As a bit of a side note, I’ve also been using Plex on my QNAP TS-653 via my Chromecast which is even more slick than Kodi so that may end up being my end solution anyway.

The thing that really got me excited about my Pi is the ability to leverage the GPIO pins to do things outside of the Pi. I’ve always had an interest in coding and have even done a bunch of Python back in the day, since this is the preferred language for those hacking on the Pi, this was quite convenient. So I ordered myself a starter kit of sorts. Don’t order that one for your Pi2 though, it comes with the wrong extension board, this one is for the original Pi. When it arrived and I realize I had ordered the wrong part, a quick trip to Lee’s Electronics and I was in business. I did what most people do and I wrote the Hello World of the Pi and hardware hacking world, I made some LEDs blink:

I continued to play with LEDs for a while, writing various python functions to make them blink in different orders but I soon grew bored and decided I needed to try another one of the external pieces of kit that I had, so I figured the 16 × 2 character display was a good next step. This was actually quite easy, so after I figured it out, I quickly whipped together some code to wish my good friend Jason a happy birthday since it happened to be the day I was working on this:

Cue the oohs and ahs!

Using a 16 × 2 character display with the RaspberryPi.

The next thing I did was attach a DS18B20 temperature sensor because now I had a project in mind. I started with some code that displayed the current temperature in Celsius and Fahrenheit to both the terminal as well as to the character display, I then added logging to it and removed the character display code. Lately I’ve been finding that my bedroom is a little warm at night, so I figured I could deploy the Pi as a temperature logging device and figure out if the room was actually getting hot or if I’m just eating too much salt or drinking too much red wine; after all, you can’t manage what you can’t measure. That’s where my Pi currently sits, powered up in the bedroom logging the temperature to a CSV file every five minutes and has been doing so for about three days. I have yet to come up with my next project, but I do know that I want/need at least one more Pi since my son Jordan has been playing with the Pi as well.  I wouldn’t want to hinder his ability to play with it and learn because I’m hogging it for the same purpose. Also, since the Pi3 is now out, I would like to have that model if only for the built-in WiFi. I’m still not done with the RetroPie project, but I am a little frustrated by it so more to come on that one in the future.

Feel free to suggest any fun projects that I may want to take on, I have a new soldering station on the way from Amazon and am looking for something cool to do with it.


Geek Defined or: How I Learned To Stop Worrying And Love So Many Things.

I’m going to start to deviate from my usual NetApp-centric posts and just talk about what I’ve been up to on the geeky side of Chris. Out of the very limited list of podcasts I listen to, one of them is The Geek Whisperers which focuses not on specific technologies but rather on personal and professional advancement from a Geek’s point of view. Due to my association with the NetApp A-Team and NetApp’s acquisition of Solid Fire, I was lucky enough to find myself on a call with the inspirational Amy Lewis, a.k.a. CommsNinja, one of the hosts of TGW. We were discussing social media, personal development and other such things. Amy made a comment about how people should be blogging about more things that they’re interested in because you never know which members of your readership may have the same interests and it may also draw in new readers.

While the term Geek used to mean either “an unfashionable or socially inept person.” or “a carnival performer who performs wild or disgusting acts.”, these are no longer the de-facto definitions. These days, Geek will often refer to someone who “engages in or discusses computer-related tasks obsessively or with great attention to technical detail” and/or some who “is or becomes extremely excited or enthusiastic about a subject, typically one of specialist or minority interest”. It is definitely this latter one that me and my geek brethren identify with. While the stereotypical “geek” typically is also a “computer nerd”, I’m betting most, if not all of them have some pretty intense interest in some non-computer related hobbies as well. For example, here’s a list of interests I’ve obsessed over in the past five to ten years, some of which are current while others have passed or are waning.

  • Photography
  • “Computers” (this one has so many facets, I almost hesitate to include on its own)
  • Sailing
  • Languages (not programming)
  • Cooking
  • RaspberryPi
  • Ingress
  • Carpentry
  • Snowboarding
  • Virtualization
  • Magic: The Gathering (This one is actually from about 20 years ago but has resurfaced now that I can play with my son)

This list is by no means exhaustive, but rather a glimpse into who I am. Personally, I find myself geeking out over so many different things, the hardest part about being a geek is finding the time to devote to such a plethora of hobbies, and then when something new comes up that strikes my fancy, finding time to fit that new one in. Maybe that’s what makes a geek a geek, someone who is willing to obsess over something that interests them until they get good at it or maybe a geek is just someone who loves to learn new things? I do find that often time my interests will somehow be related to the sciences in one way or another, be it math, physics or chemistry. I never thought I’d mention that last one until recently however when I started reading about soap making after a discussion about it with a newer co-worker; so stay tuned, there may be a soap making post in my future.

So why exactly am I posting this at all? Well I’m basically giving myself permission to post about whatever the heck I want, hopefully it’ll be enough to keep some people interested.

8.3.1 and 8.3.2…dot releases never felt so good.

NetApp released ONTAP 8.3 over a year ago now, and since then two minor releases have come as well, and with them far more payload than you’d usually expect for dot releases. Typically the major releases get all the hype, but after you see all that has been included with the two minor releases of 8.3, you’ll see what all the fuss is about.

First of all, if you can’t remember what was included with 8.3, go over here and read about it. Highlights included but weren’t limited to:

  • Metro Cluster
  • Non-disruptive LUN migration
  • Serious performance improvements in the flash space
  • Version independent SnapMirror

When 8.3.1 came out in early September, it brought some pretty spectacular:

  • More flash performance improvements
  • Storage Virtual Machine Disaster Recovery (SVM DR)
    • This is the ability to replicated entire SVMs and not just volumes to another cluster. This has two modes, Identity Preserve True or False which can replicate all the network related info for those who’s DR site supports it, i.e.: L2 connectivity.
  • In-line compression and zero elimination
  • Two node MetroCluster, i.e.: one per site
    • Uses ATTO bridge to connect the disk
    • This is more of a “Stretch MetroCluster” and is suitable for campus level DR where the loss of a building is being protected against.
  • Some performance metrics now available in System Manager


  • Copy Free Transition
    • This has got to be one of the coolest features so far, it lets you stand up a new cDOT system with minimum disk, then move your 7-mode disk over to it without having to do a data migration.
  • In-line deduplication
  • More performance improvements for SAN on AFF
  • In-place, adaptive compression
  • Fibre Channel over IP for MetroCluster
    • Up to 200km, between switches that support it, such as the Cisco 9250i
  • Quality of Service policies previously limited to 8 notes can now be applied to up to 24
  • System Manager Improvements:
    • Cluster performance charts with IOPS and latency available within System Manager
    • Manual IP assignment
      • Previously you had to create the subnet, that is no longer the case
    • SyncMirror (introduced in 8.3 with Metro Cluster) support in System Manager
      • This is not the same as synchronous SnapMirror, which is still not available in cDOT
    • You can now manage your MirrorVaults in the GUI
    • Various other System Manager improvements, far too many to list.

As you can see by the points I’ve covered off, the dot releases of cDOT 8.3 have been packing quite the payload, I’m sure that not having support 7-mode in the same release has helped speed up the development cycle for many features not to mention some of those engineers have probably been reassigned to cDOT work. I’ve left some of the more esoteric details out, but if you want to see them all, head over here to read the release notes for the individual versions.

NetApp Certifications, What’s New and What You Need to Know.

If you’re a NetApp nerd like myself or if you prefer to call yourself an “avid NetApp user”, then you’re probably familiar with their annual conference, NetApp Insight and the fact that is is just around the corner. Since you’re reading this article at all, you may already have or have at least considered getting certified. There’s not a lot new since the major update back in April when the exams were updated to reflect the release of 8.3 but there is at least one completely new exam and certification, the NetApp Certified Storage Installation Engineer, Clustered Data ONTAP NS0-180, which becomes available on September 23, 2015.

This year at Insight, there’s going to be a whopping 14 separate exam prep sessions at both the Las Vegas and Berlin versions of the conference covering the following:

  • NS0-155, NCDA 7-Mode
  • NS0-157, NCDA cDOT
  • NS0-505, NCIE-SAN E-Series
  • NS0-506, NCIE-SAN cDOT
  • NS0-511, NCIE-Data Protection

The beauty of Insight is that during the course of the conference, you can take as many exams for free as you’d like as long as it falls within their exam retake policy. If the exam centre is anything like year’s past, then it will be very busy and I highly recommend you pre-register for up to there of your exams now over here.

While we’re on the topic of certifications, NetApp is going to show the proverbial love to those of us who are already certified as well as to those who get certified while at Insight. I won’t give away all the details, but there will be different schwag based on what certifications you already hold. They’re also going to hold the first ever Appreciation and Recognition event for the NetApp Certified.

So with all this talk about certifications, lets talk about getting prepared for getting certified. The first thing you should do is follow @NetAppCertify on Twitter, join in the discussion over at the NetAppU Community and peruse the materials and sample exams available here. Sample exams are available for NS0-157, NS0-506, NS0-511 and the latest addition, NS0-180. If you’ve already have your NAIPCDOT, you’ll need to to earn the NCSIE cDOT by November 1, 2016, so I’m sure this will be a popular one. For the complete low-down on what NS0-180 might mean to you, check out this NetApp Community entry here. Lastly, be sure to check out The Value of NetApp Certification Video as well, especially since some of my friends are in it.

Finally, to further emphasize the value of NetApp Certifications, starting in October you’ll be able to add all new digital badges to your LinkedIn profile which will help job-seekers and recruiters find each other.


This new Digital Badge helps protect the value of your certification as well as providing easy verification of your NetApp Certifications.


Insight is just over a month a way as of this writing, and it’s time to start studying so that you can take advantage of those free exams which by now you’ve registered for, right? I know I have.

At the Las Vegas version of Insight, make sure you stop by The Geek & Greet Certification Appreciation Event, Wednesday at 5:15 and say hi to me and my fellow A-Team members, we may even buy you a beer.

NetApp announces Clustered Data ONTAP 8.3

Today NetApp announced the next major release of its Clustered data ONTAP operating system and a major release it is. This is the first release of ONTAP that does not include the dual payload of both 7-mode and cluster-mode and will be the norm going forward. This release has three major themes:

  1. Flash, data protection, multi tenancy, cloud, and efficiency enhancements
  2. Simplified Deployment, upgrade, transition, and support
  3. Clustered ONTAP in mission critical environments with MetroCluster

Flash, data protection, multi tenancy, cloud, and efficiency enhancements

The first theme brings with it performance enhancements in the following ways:

  • More consistent and predictable performance and higher IOPS at lower latency in the All Flash FAS (AFF) and other flash-enabled systems thanks to read-path optimization.

Random Read IO

  • The CIFS lock manager has been paralleled bringing improvements to CIFS-based file-services workloads.
  • The initial transfer as well as incremental updates for both SnapMirror and SnapVault relationships have been improved.
  • 8.3 has been optimized for more CPU cores bringing performance enhancements to pre-FAS8000 systems. Initial claims are that FAS62xx performance is similar those running 8.1 while the FAS3xxxx and FAS22xx are showing 8.1-type performance in SAN deployments.

As far as efficiency enhancements are concerned, a long awaited feature by myself is Advanced Disk Partitioning (ADP) which has three use cases:

  1. Root-data partitioning for All Flash FAS (AFF) systems.
  2. Root-data partitioning for Entry-level platforms.
  3. SSD partitioning for Flash Pools

The first two use cases mentioned above will greatly ease the dedicated root aggregate disk tax which has been the bane of the SMB buyers since cDOT’s initial (non-GX) release, providing 20+% increase in storage efficiency in 24-drive FAS255x as well as the FAS2240. This will be the default configuration for systems purchased with 8.3 but if you wish to retrofit an existing system you’ll have to evacuate your data and start fresh. As far as the third use case is concerned, the benefit here is the parity disk tax as represented by the graphic below:



Other efficiency enhancements come in the way of addressable cache, in fact the complete complement of contemporary systems (read: FAS80xx and FAS25xx) has been quadrupled. Also, the 16KB cutoff for Flash Pool has been eliminated, compress blocks are now read cacheable as are read-only volumes such as SnapMirror and SnapVault destinations.

Simplified Deployment, upgrade, transition, and support

In the never ending quest to make their product easier to deploy, transition to and use NetApp brings the following laundry list of improvements.

  1. System Setup 3.0
    • Support of AFF aggregate creation
    • 8.3 networking support (More on this in a subsequent post.)
    • Four port cluster interconnect support
  2. System Manager 3.2
    • This becomes a cluster-hosted web service which can be reached from the network using Mozilla, Chrome and IE on Windows, Linux and Mac platforms.
    • 8.3 networking support
  3. Automated NDU
    • Three commands to upgrade your cluster.
    • One command to monitor the progress.
  4. Networking
    • There is a whole litany of changes/improvements, too many to list here. The biggest one however may be IPSpaces so know you can have overlapping subnets in those multi-tenant environments.
  5. Virtualization
    • vVol support (pending VMware support)
    • FlexClone for SVI
    • Inline zero write detection and elimination.
  6. 7MTT
    • Version 1.4 will bring with it a new collect and asses feature to validate the destination cluster based on the assessment of the source 7-mode system.
    • 2.0 brings with it the much sought after SAN migration.

Clustered ONTAP in mission critical environments with MetroCluster

Not a whole lot more to say around that except that it is finally here. Some of the highlights are:

  • Two node cluster at either site
  • Clients can be served from all four nodes at the same time
  • Support for Non Disruptive Operations (NDO)

While I covered a lot in this post, I didn’t cover everything as 8.3 is a major release indeed. Now the big question many of you will have is what platforms will support it? Look no further:

  • FAS8xxx
  • FAS25xx
  • FAS62xx
  • FAS32xx (except the FAS3210)
  • FAS22xx

As for what I didn’t cover in this post but you may wish to research further:

  • VM Granular Management
  • 8.3 style networking
  • DataMotion for LUNs
  • Offline Foreign LUN Import
  • Version Independent SnapMirror (this one’s pretty cool)
  • Other Performance Improvements
  • Further Protocol Enhancements (SAN and NAS)
  • Data ONTAP in the cloud (Cloud ONTAP)

NetApp Refreshes Entry-Level FAS Systems

Today NetApp announced the successors to their entry-level line of FAS storage arrays: the FAS2552, FAS2254 and the FAS2520 which replace the FAS2240-2, FAS2240-4 and FAS2220 respectively.

Why is this important? Until now, in order to run Clustered Data ONTAP, you had to use your one and only expansion option for a 10GbE card for the cluster interconnect network, giving up any chance of deploying Fibre Channel. Technically, since this was a two-port card, you could still provide 10GbE uplink at the expense of redundancy on the ClusterNet backend. However, the new models give up the mezzanine slot altogether in favour of a minimum of 4 ×10GbE on board on the FAS2520 to 4 ×UTA2 ports on both the FAS2552 and FAS2554.


With this refresh NetApp continues to use the same dual-core, hyper-threaded, 1.73GHz Jasper Forest processors as before – which, incidentally, was specifically designed for both embedded and storage applications — but the quantity is doubled to four, not to mention there’s a three-fold increase in memory. All of this added memory increases the ability for Data ONTAP to address more flash, raising the Flash Pool™caching limit to 4TB. Finally, with the addition of onboard 10GbE across the line, NetApp closes the gap in regard to ClusterNet interconnect requirements. The minimum version of ONTAP required for either 7-mode or Cluster-Mode will be the one it ships with, 8.2.2RC1.


The FAS2520A is a 2U appliance supporting 12 SAS, SATA, and NSE drives internally, and up to 72 additional drives externally. Connectivity is provided by 4×6GB SAS ports, 4×1GbE interfaces and 8×10GBASE-T. Unlike its predecessor, there are no expansion slots.


NetApp’s new FAS2520, rear view.


The FAS2552A is a 2U appliance supporting 24 SAS, NSE and SSD drives internally and the FAS2554A is a 4U appliance supporting SATA, NSE and SSD drives internally; both models support up to an additional 120 drives externally. Connectivity is provided by 4×6GB SAS ports, 4x1GbE interfaces and 8×UTA2 ports. The UTA2 ports can be configured as either 8Gb FC, 16Gb FC, or 10GbE. The 10GbE configuration does indeed support FCoE as well as the usual CIFS, NFS and iSCSI options. Due to the fact that each pair of ports is driven by one ASIC, the UTA2 ports must be configured in pairs. However, it should be noted that their personality can be modified in the field; this requires a reboot as well as the requisite SFP.


NetApp’s new FAS2552, rear view.


NetApp’s new FAS2554, rear view.

Port Legend


With this second round of major updates to the FAS systems this year, the entire line is now truly Clustered Data ONTAP-ready, with every model sporting 10 Gig connectivity on-board. What I find most noteworthy is the amount of RAM that has been added which significantly increases the amount of flash-based cache the devices can address. Flash Pools abound!