Archives for June 2016

Nutanix AOS 4.7 Immediately Available

What’s New in AOS 4.7? Some of these enhancements were announced at .Next last week. Here’s a full list of what’s new in AOS 4.7. It is available for immediate download on the Nutanix support portal.

Acropolis Block Services

Acropolis Block Services (ABS) provides highly available and high performance block storage as iSCSI LUNs to clients. Clients can be non-Nutanix servers external to the cluster or guest VMs internal or external to the cluster, with the cluster block storage configured as one or more volume groups. This block storage acts as the iSCSI target for client Windows or Linux operating systems running on a bare metal server or as guest VMs using iSCSI initiators from within the client operating systems.

Cluster Resiliency (Degraded Node Detection)

Nutanix Engineering has introduced capabilities in AOS to identify partially available nodes and prevent them from affecting cluster availability. It can detect if a node is experiencing  network latency issues, disk or memory errors, or other issue that would affect cluster performance. AOS marks the node as degraded, places it in maintenance mode, and raises an alert to indicate its status. The node will be prevented from participating in cluster operations and cluster operations will continue. The node can rejoin the cluster once the issue is fixed. This feature is disabled by default and customers are advised to contact Nutanix Support to enable and use it. Reference internal Knowledge Base Article 3361.

Docker Container Support

The Nutanix volume plug-in for Docker enables Docker deployments to be integrated with external storage systems, in this case the Nutanix distributed file system (NDFS), and enable data volumes to persist beyond the lifetime of a single container machine host.

1-click HBA Disk Controller Firmware

The LSI host bus adapter/controller (HBA) 1-click firmware upgrade feature is available for Acropolis hypervisor (AHV) and ESXi hypervisor host environments running on NXxxxx G4 (Haswell) or NX-xxxx-G5 (Broadwell) platforms only.

In-Place Hypervisor Conversion

Previously tech preview, this 1-click feature available through the Prism web console allows you to convert your cluster from using ESXi hosts to using AHV hosts. Guest VMs are converted to the hypervisor target format, and cluster network configurations are stored and then restored as part of the conversion process. Note: This feature converts your existing ESXi cluster to an AHV cluster. You cannot start the conversion process on the AHV cluster.

Intel Broadwell Architecture

Support for platforms using the Intel Broadwell CPU and architecture.

Internationalization (i18n)

Non-ASCII user defined values (for example, entity names, descriptions, and so on) are supported in UTF-8. For the list of supported and unsupported entities and to change language settings to Simplified Chinese, see the Prism Web Console Guide.

Newly-supported platforms

Support for Nutanix Xpress platforms such as the SX-1065-G5. See the Xpress Series Compatibility Matrix in Release Notes | NX-series Hardware on page 22. Support for Lenovo Converged HX Series

Prism Central

Enhanced Prism Central Explore tab usability with custom page sizing, custom Focus views, and easier select options.


Nutanix Security Engineering and Research Team (nSERT) has identified existing vulnerabilities associated with certain security concerns and this release addresses these vulnerabilities. See Resolved Issues | AOS 4.7 on page 6 of Release Notes for details.

Simplified SSD breakfix procedures for Hyper-V

Support for single SSD replacement procedures for Hyper-V clusters.

Support for the Microsoft Cloud Platform System (CPS)

Support for the Microsoft Cloud Platform System (CPS), which bundles Windows Server 2012 R2, System Center 2012 R2 and Windows Azure Pack for easier hybrid cloud configuration. Nutanix offers the CPS Standard version preinstalled on nodes from the factory.

TRIM Support for Acropolis File Services

For Acropolis File Services, the Prism Web Console previously could not show accurate storage consumption as it did not show deleted data. Now, Acropolis File Services introduces TRIM and thin provisioning with 4.7 for better storage. TRIM is enabled by default and communicates free block information to the Prism Web Console from TRIM. AOS then performs storage distribution for better storage consumption.

Whitelist Prism Central and Prism Element to bypass proxy servers

Prism Central and AOS 4.6.2 introduce new nCLI options ncli http-proxy add-to-whitelist / ncli http-proxy delete-from-whitelist. Use the ncli http-proxy add-to-whitelist command to bypass a proxy server used by a Prism Central-managed cluster and allow network traffic between Prism Central and the cluster. Previously, if you attempted to register a cluster that implemented a proxy server, the registration failed. The Prism Central Guide describes how to configure this use case with these new commands.

.Next 2016: Don’t leave behind Baremetal Workloads

Presenters: Mike McGhee (NTX), Tabrez Memon (NTX), Tim Issaacs (NTX)

Why bare metal? Oracle Licensing, performance perception, Legacy apps (Sun, IBM, HPUX), investment protection

High Performance Databases – a few times you might not virtualize

  • Licensing cost
  • Virtualized limits for max DB (SAP HANA, etc.)

3-tier applications

  •  Run your 3-tier app entirely on Nutanix
  • Nutanix can run VMs plus block storage

How ABS (Acropolis block services) works:

  • Serve block storage via iSCSI to physical servers
  • Eliminates the need for iSCSI MPIO
  • Client intervention not required when the cluster scales out
  • Inherits all Nutanix features
  • Run VMs while serving external storage

First release RHEL 6, Windows Server 2012 R2, Oracle 6. On the road map is support for other OSes including ESXi (~3 months).

iSCSI supports SCSI-3 persistent reservations for clustered volumes

Evenly balances I/O across all cluster nodes

Tabrez goes into a deep dive of how we negate the need for client MPIO. Basically we hide all the complexity using various iSCSI commands like virtual targets, iSCSI redirection, etc.

You can create a ABS sub-group, and limit the iSCSI connections to a subset of the cluster nodes.

All data reduction features are supported (compression, dedupe, ECX), thin provisioning, SCSI UNMAP, CHAP authentication

Use cases:

  • Exchange + iSCSI Clustering
  • Shared storage clustering
  • MS SQL Clustering
  • Oracle RAC
  • Bare metal
  • scale-up performance

.Next 2016: Business Continuity for Tier-1 Apps

Presenters: Partha R (NTX), Mike McGhee (NTX), Ryan Sheldon (AMGEN)

Tier-1 apps are different for different customers. Could be ERP, EPOS, email, etc.

Need to be prepared for the worst: IT Productivity, recovery, end-user productivity, business disruption, lost revenue, etc.

Data protection requirements: RPO, RTO

Understand consistency: Application consistency (e.g. VSS). Most beneficial in the context of backup and restore.

Technologies: Application (Exchange DAGs, SQL AAGs, Oracle Dataguard), Hypervisor (hyper-v replica, vsphere replication), storage (infrastructure centric, Nutanix Data Protection)

What am I protecting? Application dependencies? One or multiple methods? How frequently do you backup and replicate? Sync vs. async?

Where does Nutanix native replication fit in?

Timestream – RTO of minutes, RPO of minutes

Cloudconnect – RTO hours, RPO hours

Async replication – RTO minutes, RPO minutes

Sync replication – RTO near-zero, RPO zero

Nutanix Local Snapshot (Time stream)

  • Protection against guest os corruption
  • Snapshot of VM environments
  • VM or vdisk granularity
  • Low performance impact
  • VM and application level consistency

Nutanix Async Replication

  • Delta changes
  • Dedupe on the wire
  • Compression on the wire
  • Flexible topologies
  • Bandwidth schedules

Nutanix Cloud Connect

  • Hybrid cloud solution from Nutanix
  • Integration with Azure and AWS
  • For archiving and backup
  • Easy to setup and manage
  • WAN optimized replication
  • Interop with Nutanix DR

Volume Groups with Async Replication

  • Exchange + iSCSI
  • MS SQL Clustering
  • Oracle RAC
  • Bare Metal

Consistency options: VSS

  • VMware Tools or Nutanix guest tools
  • Pre and post scripts for Linux

Consistency options: Consistency Groups

  • For a group of VMs and a consistent restart point


Manufacturing is key. They leverage Metro availability to provide high uptime. Customer set it up in about 10 minutes. With MA they can converge services like DNS, DHCP, etc. and save on licensing costs.


What’s next: Metro Availability Witness coming out in Asterisk, which will automate site-failover.

.Next 2016: Day 1 First General Session

Here we are, at the second Nutanix .Next conference! Last year it was in beautiful Miami, and this year it’s in hot Las Vegas. I’m not sure which is hotter, the weather or our announcements. Like last year, this event is sold out and more than double the size! I’ll be live blogging the sessions that I’m attending this year. As always, then are real-time notes and don’t have time to take slide photos and insert them in the blog posts. But I try and capture the essence of each session

All of the attending Nutanix NTCs are lined up on stage and several of them do introductions.

Howard Ting takes stage.

In our first year we expected 500 to 700 attendees, we ended up over 960 people. This year the goal was 2000, and our registration total is 2,500.

Howard covers some of our great customers, such as airlines that run critical airline operations, a company that prints billions of coupons on Nutanix, or hospitals running critical operations on Nutanix. “You’ll love Nutanix”. “Love” is not a word you hear much about IT infrastructure. Nutanix is on a journey to build the enterprise cloud. In reality, only 7% of companies have built a true private cloud. Most think they have just by virtualizing.

Nutanix: Enterprise  Cloud Platform

Dheeraj Pandey (CEO/Founder) takes the stage

The Growth Paradox: Most executives cite internal factors, not market forces, for what slows growth.

Founder’s mentality: Insurgent mission, owner mindset, obsession with the front line

Ambition, Attention-to-detail

One click upgrades: One click mover

Dheeraj talks about Community Edition and brings Richardarsenian on stage. They show off a drone which has Nutanix community edition running and controlling it.

DataGravity: Data driven security

“Coming together is a beginning; keeping together is progress; working together is success.”

Sunil Potti takes the stage

Nutanix wants to be the AWS for the enterprise – Enterprise cloud platform on my terms

Shipping very shortly  will be v4.7 and Asterix

Within one week of the 4.6 release, 500 clusters were upgraded. After 100 days, nearly 50% of all clusters were upgraded.

We support: Mainstream virtual apps, now all virtual apps,

Sunil shows a performance graph, where customers got a 4x improvement for free just via software upgrades.

For Broadwell servers all models now support all flash configurations. Or hybrid, your choice.

Michael Tempora from Vonage comes on stage. They are rapidly expanding into the business VoIP sector. Vonage cloud platform is powered by Nutanix.

Why stop a virtual workloads? Nutanix now supports bare metal via Acropolis Block Services featured in 4.7. Run 3 tier apps entirely on Nutanix. 200,000 random Oracle IOs in 2U.

Demos Citrix on AHV with MCS integration and Acropolis File Services (AFS). They also show the cluster running heavy SQL loads, plus bare metal Oracle workloads using ABS. The true power of a unified fabric.

15% of Nutanix clusters now running AHV

Jake Yang, Nasdaq Global Infrastructure director. Believes the hypervisor is now a commodity. Big adoption of AHV.

New: Acropolis Container Services. A native container platform, Docker support, persistent portable storage volumes.

One click upgrades: BIOS, BMC, Storage controller, disks, hypervisor, AOS, NCC, Foundation

Asterisk: 1-click self-service. Shows a demo of the new “Self-Service” tab in PRISM.

On demand Nutanix Community Edition: Can launch a public cloud hosted CE cluster and use for free for 2 hours.

Asterisk: You will no longer need to login to vCenter to manage ESXi. PRISM will support multi-hypervisor support.

T-Mobile’s Kevin Luu, Sr. Director Infrastructure and DevOps to talk Nutanix.

AOS 4.7: Microsoft CPS on Nutanix. Factory installed, deployed in hours, 1 click upgrades, full stack support from Nutanix.

Clutter-free organization: New filtering, labels, and great new search features

Machine intelligence: X Fit algorithms, Predictive workload scheduler, just in time forecasting. Recommends new hardware based on predicted capability.

Asterisk: Visualization of the network. An all new network view coming to PRISM.