Behind the Scenes: Episode 147 – SPC-1v3 Results – NetApp AFF A800

Welcome to the Episode 147, part of the continuing series called “Behind the Scenes of the NetApp Tech ONTAP Podcast.”


This week on the podcast, we find out how the new NetApp A800 system fared in the rigorous SPC-1 v3 storage benchmarks. Can the NVMe attached SSDs truly help reduce latency while maintaining high number of IOPs? Performance TME Dan Isaacs (@danisaacs) and the workload engineering team of Scott Lane, Jim Laing and Joe Scott join us to discuss! 

Check out the published results here:

And the official NetApp blog:

Finding the Podcast

The podcast is all finished and up for listening. You can find it on iTunes or SoundCloud or by going to

This week’s episode is here:

Also, if you don’t like using iTunes or SoundCloud, we just added the podcast to Stitcher.

I also recently got asked how to leverage RSS for the podcast. You can do that here:

Our YouTube channel (episodes uploaded sporadically) is here:

New NetApp FlexPod Cisco Validated Designs available!


In case you weren’t aware, NetApp partners with other vendors, such as Cisco, to create bundled architectures that are co-supported and simple to deploy. These converged infrastructure products are known as “FlexPods” and they go through a rigorous testing and validation process by both NetApp and Cisco to ensure the highest probability of success in deployments.

There are two new Cisco Validated Designs (CVDs) available. I’ll basically be plagiarizing the announcement emails that went out here, because I like converging communications, too. 🙂

FlexPod Datacenter (AFF A300) with Citrix XenDesktop/XenApp for 6000 Seats (credit to Chris Gebhardt – @chrisgeb)

This CVD is a new milestone in user density on the A300 platform running ONTAP 9.3 at 6000 users.  In this CVD we were able to observe peak total values of 83,000 IOPS and a bandwidth of 1750 MB/s at a latency of less than 1ms.  This low latency translates directly into an excellent end-user experience on mixed use case scale testing scenario.

The solution provides customers with a blueprint for deploying large scale mixed use case of persistent, nonpersistent & RDS Windows 10 virtual machines on FlexPod.  Cisco & NetApp have stress-tested the limits of the infrastructure during many different scenarios.

Some of the key values of this solution include:

  • NetApp AFF A300array provides industry-leading storage solutions that efficiently handle the most demanding I/O bursts (for example, login storms), profile management, and user data management, deliver simple and flexible business continuance, and help reduce storage cost per desktop.
  • NetApp AFF A300array provides a simple to understand storage architecture for hosting all user data components (VMs, profiles, user data) on the same storage array.
  • NetApp ONTAP 9.3 softwareenables to seamlessly add, upgrade or remove storage from the infrastructure to meet the needs of the virtual desktops.
  • NetApp Virtual Storage Console(VSC) for VMware vSphere hypervisor has deep integrations with vSphere, providing easy-button automation for key storage tasks such as storage repository provisioning, storage resize, data deduplication, directly from vCenter.

 More information on this and other solutions can be found:

Cisco NetApp Website:

All Flash FAS Datasheet:

Additional FlexPod solutions with All Flash FAS, AltaVault, SAP HANA, Microsoft SharePoint, VMware vSphere, Cisco UCS Director and more, can be found:

FlexPod Datacenter with VMware 6.5 Update1 and Cisco ACI 3.1 Solution (credit to Dave Derry)

I am pleased to announce the release of an updated FlexPod Solution –  FlexPod Datacenter with VMware 6.5 Update1 and Cisco ACI 3.1.  This validated FlexPod design (CVD) refreshes FlexPod with All Flash FAS with ONTAP 9.3, Cisco ACI 3.1, and VMware vSphere 6.5 u1.


This solution continues the momentum of the huge success of NetApp All Flash FAS along with over 5000 Cisco Nexus 9k customers and over 200 production deployments of Cisco ACI in enterprises and service providers.

FlexPod with All Flash FAS and Cisco ACI is the optimal shared infrastructure to deploy a variety of workloads.  FlexPod provides a platform that is both flexible and scalable for multiple use cases and applications.   From virtual desktop infrastructure to SAP®, FlexPod can efficiently and effectively support business-critical applications running simultaneously from the same shared infrastructure. The flexibility and scalability of FlexPod also enable customers to create a right-sized infrastructure that can grow with and adapt to their evolving business requirements.

FlexPod is the leading converged infrastructure with the lowest storage cost per TB of any converged infrastructure in the market, according to IDC.  This coupled with All Flash FAS increases application performance by 20X and with Cisco ACI that provides up to 83% faster network provisioning times for applications compared to existing network provisioning tools.

Use Case Summary

This infrastructure solution validates this design including redundancy, failure and recovery testing:

  • UCS B200-M4 and M5 compute servers, 6300-series fabric interconnects, running UCS Manager 3.2
  • Nexus N9k running in ACI Fabric Mode, 40G end-to-end
  • Cisco APIC managing the Nexus 9k fabric, running ACI 3.1
  • AFF A300 controllers with ONTAP 9.3, with storage connected to the ACI fabric (NFS)
  • OnCommand System Manager, OnCommand Unified Manager, OnCommand Workflow Automation, NetApp Virtual Storage Console for vSphere
  • NetApp SnapDrive and SnapManager
  • VMware vSphere 6.5 u1
  • Validation of FC direct-connect and iSCSI for SAN boot and data access from the fabric interconnects to storage
  • FC, iSCSI, and NFS Storage access
  • NetApp storage QoS to limit load to keep CPU utilization at a recommended level

The FlexPod Datacenter with VMware 6.5 Update1 and Cisco ACI 3.1 design guide is located here:

To review additional validated FlexPod solutions including validated infrastructure and workload solutions, see this link:

We also will have a podcast soon that covers the CVD for OpenStack in a FlexPod. Stay tuned for that!

The NetApp A200 is kind of a big deal…


Recently, reviewed the NetApp All Flash FAS model, the A200.

The testing workload was as follows:

The application workload benchmarks for the NetApp AFF A200 consist of the MySQL OLTP performance via SysBench and Microsoft SQL Server OLTP performance with a simulated TPC-C workload.

Some of the application simulated workloads were:

  • VMware
  • Oracle
  • SQL

Overall, the performance seen from the StorageReview testing was well received, especially considering the fact the A200 is an entry-level flash system. The A200 blends a nice mix of performance, capacity (over 360TB raw) and density (2U of rack space!) in a single system. As a bonus, the raw capacity is measured *before* the 4:1 storage efficiency guarantees provided by NetApp.

As a result, tagged the A200 as an editor’s choice. For the full review, click on the link below:

Behind the Scenes: Episode 71 – SPC-1 benchmark and the A700s

Welcome to the Episode 71, part of the continuing series called “Behind the Scenes of the NetApp Tech ONTAP Podcast.”


We decided to do a mid-week release of the podcast to talk about the brand new AFF platform, as well as the insanely high results from the SPC-1 benchmark test! We bring in the performance team, featuring Chad Morgenstern and Jim Laing, to talk numbers. Then, we invite Mr. Flash, Andy Grimes (@andy_ntap_flash),  to give us the lowdown on the new compact beast, the A700s.

Finding the Podcast

The podcast is all finished and up for listening. You can find it on iTunes or SoundCloud or by going to

Also, if you don’t like using iTunes or SoundCloud, we just added the podcast to Stitcher.

I also recently got asked how to leverage RSS for the podcast. You can do that here:

You can listen here:

TECH::Smaller is better; less is more. – Introducing the new 6U 8080 AFF!


If you’ve ever run a data center, you know that physical rack space is at a premium. I’ve been reminded of this recently, as me and several of my TME cohorts have been busy working in the lab getting something special ready for this year’s NetApp Insight. I’ve been tweeting videos of it as teasers and will have a new one coming out this week.

Basically, we’ve been trying to squeeze an entire demo into a travel rack. While, this isn’t something most people would try, space for physical servers and storage is a concern for every datacenter administrator, especially service providers.

Enter AFF

NetApp recently has been aggressively pursuing the all-flash market with some super deals for customers. They are outlined in various places:

There was also a recent Tech ONTAP podcast on Flash:

But the problem is, one of our beefiest models – the AFF8080EX – was a taking up a whopping 12U of rack space! That needed to get better, and it has.

Smaller is better, less is more

The Register reported on this as well, but my goal is to explain why a smaller footprint of rack space is better for any data center admin, if it’s not already apparent.

Data center real estate is a real cost that many people forget when looking at the overall price tag for servers, storage, switches, etc. The reason it gets overlooked is that it’s a price tag that accumulates over time and isn’t readily apparent up front. Data center costs are broken down into a few areas.

Physical rack space: The more rack space you eat up, the more racks you have to buy. More racks = less physical floor space. That means you end up having to rent or buy more space.

Power: Bigger units generally eat up more power (not always, but generally). That means you pay more per month in electricity.

Cooling: Bigger units generally require more BTUs for cooling. This also eats more power.

Using more dense equipment saves money in the long run and it’s especially useful when those servers/storage that are more dense also have better performance than servers/storage with larger footprints.

It’s just one more way NetApp AFF is making all-flash storage more of a reality for the data center.

TECH::Clustered Data ONTAP 8.3.1 is now in general availability (GA)!

Looking for cDOT 8.3.2? Check it out here:

cDOT 8.3.2 is the first release that offers Copy-Free Transition!

UPDATE: Back in June, clustered Data ONTAP (cDOT) 8.3.1 became available as a release candidate (RC). (I cover what a release candidate is in my “What’s the latest 8.3 release?” blog.)

Now, it’s available in General Availability! That means, have at it!

NOTE: Be sure to get the latest patch release of 8.3.1, which is currently:

I also cover the latest All Flash FAS promotional updates in “NetApp is kicking some flash!


Despite having the designation of being a “minor version,” this release brings new features to clustered Data ONTAP, which is why it’s getting RC designation. If you’re running clustered Data ONTAP (especially for NAS environments), this is the version to be on, hands down.

Each section includes links to documentation. Some of the docs might not be updated until 8.3.1 goes to General Availability (GA), so keep checking back!

New features


The 8.3.1 release brings a number of new features that will greatly improve performance, transition to clustered Data ONTAP and overall cluster resiliency.

Improved inline data compression

Data compression in clustered Data ONTAP allows for greater storage efficiency by compressing data within a FlexVol on primary, secondary and archive storage.

Some of the improvements include:

  • Support for all workload environments
  • Optimization for All Flash FAS (AFF) systems (where compression is enabled by default)
  • Adaptive compression
  • Improved read performance
  • Sub-file clone support on compressed volumes

For more information regarding compression in 8.3.1, see the product documentation.

SnapVault support for inline adaptive data compression

In cDOT 8.3.1, you can leverage the new inline adaptive data compression! (Provided both source and destination support it)

For more information on this, see TR-4183.

Foreign LUN Import Improvements

cDOT 8.3.1 brings two major improvements to Foreign LUN Import (FLI), which greatly enhances the transition story:

  • Support for ONLINE FLI!
  • FLI Throttling

This includes the ability to import LUNs from 7-Mode to a cDOT cluster, making transition *that* much easier!

For more information regarding Foreign LUN import in 8.3.1, see the product documentation.

Also, check out the new FLI TR-4442!

2-Node MetroCluster (MCC)

Prior to clustered Data ONTAP 8.3.1, MetroCluster was only supported on 4 nodes. cDOT 8.3.1 introduces support for MCC configurations of 2 nodes, with a single node cluster at each site.

For more information on MetroCluster in cDOT, see TR-4375.

Security Audit Log Forwarding

cDOT 8.3.1 now allows forwarding the command-history logs to a remote server. This allows for flexibility and security management for storage administrators.

For more information on events in cDOT, see TR-4303.

Support for cluster peering (SnapMirror) in non-Default IP Spaces

I cover what IP Spaces and Broadcast Domains in cDOT are in my DataCenterDude blog post. This new feature in cDOT 8.3.1 allows SnapMirror relationships to use independent IP spaces for SnapMirror. The full-mesh network requirement seen in previous versions of cDOT is now only required at the IP Space level.

For more information on networking in cDOT, see TR-4182.

NAS Improvements

In cDOT 8.3.1, NAS has vastly improved. In fact, it is the recommended clustered Data ONTAP version for all NAS environments. These include:

  • Ability to modify credential cache timeouts
  • New options including cifs.nfs_root_ignore_acl, nfs.ntacl_display_permissive_perms
  • SMBv3 Encryption Support
  • Better export policy rule cache handling
  • Better netgroup cache handling
  • Support for Windows NFS (previously not supported in any 8.3.x release)

For more information on NAS improvements in 8.3.1, see TR-4067 and {need CIFS/SMB TR}

And last, but certainly not least…

Storage Virtual Machine Disaster Recovery!

That’s right! The analog to vfiler DR is now available in cDOT 8.3.1. You can replicate entire SVM configurations and data to remote sites and failover when disaster strikes. For more information on SVM DR, see my blog post SVM DR in cDOT!

Supported platforms

The following shows supported platforms for 8.3.1.

  • FAS2xx0: FAS2220, FAS2240
  • FAS/V 3xx0: FAS/V 32×0 all models except 3210
  • FAS/V 6xx0: FAS/V 62×0 all models
  • FAS 80xx: 8080, 8060, 8040, 8020
  • FAS 25xx: 2554, 2552, 2520

Systems that are NOT supported with AFF in 8.3.1:

  • FAS/V 3xx0: 31×0 all models, 3210
  • FAS/V 6xx0: 6040, 6080
  • IBM n-Series

TECH:: NetApp is kicking some flash!

Picture this…

You’re a storage administrator, and your boss has just told you to go out and find a suitable flash array for your production workloads. The problem is, everything you’ve heard about flash is that there isn’t a flash array that can do unified protocols (SAN and NAS) *and* deliver the top-line performance with built-in data protection at the lowest cost per GB out there.

That changes today.


Customers and partners alike have been asking for an affordable, dependable and reliable high performance, enterprise-ready all flash system – and NetApp listened.

Recently, there has been a lot of talk about flash replacing disk as a primary storage mechanism in a datacenter. Because of the price point, performance and other factors, flash storage has become coveted in IT organizations.

There are plenty of storage vendors out there, but only one has the proven track record in enterprise environments and the ability to offer unified protocol support, flash-to-disk-to-cloud, non-disruptive operations (NDO), and integrated data protection on a single storage system.

NetApp All Flash FAS is stepping up its game with a new, aggressive pricing model that allows enterprise IT departments to implement a scale-out All Flash storage system without needing to sell a kidney to do it.

Now, you can get:

  • Consistent high performance at a low latency
  • Ability to scale up to 4 million IOPS and 16PB of effective capacity
  • The enhanced features, non-disruptive operations and flexibility of clustered Data ONTAP
  • The only all flash array that supports combining hybrid and all flash systems into a unified storage resource!

Best of all, it’s very competitive in the industry – as low as $5/GB!

This includes software license bundles, support, and a 3 year basic warranty. So, for competitive, if not superior pricing, you get world-class, enterprise-ready flash storage from the only vendor that can offer enterprise-ready flash storage!

Think that’ll make your boss happy?


Just think about how happy he/she will be when you tell them that you bought an all flash system that can also incorporate spinning disk for all those workloads that don’t need to live on flash, such as archives, home directories and other things that aren’t so performance hungry.

Because, let’s be honest with ourselves – you don’t need flash for everything.

If anyone tells you otherwise, they’re being purely dishonest.

That same unified storage system that has the performance and efficiency benefits of flash, also has the ability to do non-disruptive moves of data to spinning disk when those performance needs are filled.

Have a seasonal workload that needs top line performance for a short period of time, but doesn’t do anything the rest of the year? Use your All Flash FAS nodes for the performance and then use volume move to migrate the data to spinning disk, freeing up valuable flash storage real estate for your needy workloads.

That gives you a storage system that is ideal for any and all workloads. Want to do VDI, Oracle databases and SQL on flash and archives, home directories and other capacity-based workloads in the same namespace?

NetApp is the only vendor that offers that option.

What about 8.3.1? Does that get me anything?

Heck yes!


With the new inline data compression enhancements in clustered Data ONTAP 8.3.1, you get:

  • Support for all workload environments
  • Performance and efficiency optimization for All Flash FAS (AFF) systems (where compression is enabled by default) known as “FlashEssentials”
  • Adaptive compression 
  • Improved read performance
  • Sub-file clone support on compressed volumes

In fact, in some workload scenarios, using inline compression in clustered Data ONTAP 8.3.1 with AFF brings better performance than not using inline compression in cDOT 8.3!

That’s right! You get faster results using compression than not using it. Faster performance and lower capacity? Yes, please!

For more information on 8.3.1 features, check out my post on specific enhancements in clustered Data ONTAP 8.3.1.

Is it cloud ready?

You’re in luck – clustered Data ONTAP is ideal for the cloud.

Got a private cloud? Cool. Use your volume moves to non-disruptively move data between flash and non-flash storage.

Want to move data to a public cloud like AWS? Awesome. NetApp allows you to replicate and manage your data in the cloud via Cloud ONTAP and NetApp Private Storage.

Want to tie it all together and leverage a hybrid cloud without any constraints across your choice of resources? Then check out the NetApp Data Fabric.

NetApp’s All Flash FAS solution is the only all flash product that provides full flash-to-disk-to-cloud data management within the NetApp Data Fabric.

Still not sure?

No problem. NetApp offers a “Try and Buy” program, where you can get an All Flash FAS system into your datacenter to test various workloads and see the performance gains and advantages with your own eyes!


My NetApp AFF game face

For some more information on the NetApp Data Fabric, check out Jarett Klum’s blog!

For the Reg article on this offering see NetApp cackles as cheaper FlashRay lurches out of the door.

For Dmitris Krekoukias’s excellent take on this, see NetApp Enterprise Grade Flash.

For ESG’s Lab Review of AFF and 8.3.1 see ESG Lab Review: NetApp Clustered Data ONTAP 8.3.1 and All Flash FAS AFF8080 EX