Nutanix on the way to make the datacenter invisible!
During the .NEXT event in Las Vegas last June, Nutanix gave a great insight into the roadmap of their product for the upcoming year. At this moment, the .NEXT Europe is hosted in Vienna. During the keynote, presented on Wednesday November 9th, new announcements have been made.
Also other improvements, mentioned in Break-out sessions are included in this post.
On earlier occasions storage and the hypervisor have been made invisible by Nutanix. Next to be integrated in Prism is the Networking. Network visibility was already announced, but network orchestration will also be included. This will work with network equipment from vendors like Mellanox, Arista and Plexxi.
Vlans can be provisioned in Prism and upon creation, the switch ports will be configured using the common api’s to the switch.
Additional networking equipment like load balancers and firewalls can be easily inserted for a safe and well performing infrastructure.
Already three hypervisors are supported within the Nutanix platform. A fourth hypervisor is added. Citrix Xenserver will be available for a complete Citrix integration. There is also complete support for Citrix solutions like XenDesktop and Xenapp on AHV.
The Nutanix platform is now fully supported by Oracle. This is valid for Oracle VMs on the hypervisor AHV, as well for the storage provided by Acropolis Block Services (ABS). Impact of Oracle licensing for your deployment is not changed, so it seems that ABS is still the most cost-effective option for Oracle deployments.
Acropolis File Services (AFS)
For 9 months now, the Acropolis File Services are in Tech Preview. Finally, with the release of AOS5.0 (Asterix) AFS will be GA when running AHV and ESXi as hypervisor. For delivering File Services in your company or to your customers, you can now enjoy the known aspects of a Nutanix infrastructure. This includes scalability, availability and simplicity.
For backup purposes at least 1 snapshot a day is created. This snapshots can integrate with Windows Previous Versions, so the end user has the ability to restore his own files, if needed.
For DR purposes, native Nutanix replication is available for AFS. Also, just like for VMs, AFS CloudConnect is added, so your data can be stored safely in AWS or Azure for back-up purposes.
Quota’s can be set using Active Directory groups, or single user accounts.
It was a long wait for GA of the File Services, but the solution is very mature, especially for a 1.0 version.
The Nutanix hardware is normally shipped, based on SuperMicro hardware. Nutanix has also agreements with Dell and Lenovo for delivering the Nutanix platform on their hardware.
Lately, Nutanix and Cisco announced support for a few models of the UCS rack mounted servers. Today, at the .NEXT is announced that also UCS blade servers are supported. It will be possible to create a cluster with both types combined.
Self-Service Portal en Role Based Access
Finally, it will be possible to delegate control to users. Until now, it is only possible to give an user administrative rights or view-only permissions. With release 5.0 of Acropolis, it is possible to create user roles with specific rights on a vm. Can you very helpful to give suppliers console access to a certain VM, and maybe the possibility of restarting this VM.
Role Based Access is needed for implementing a self service portal. It will be possible to create projects with a certain amount of vCPUs, RAM and storage. Users assigned to this project can deploy as much VMs as the needed, as long the resources of the project are not used completely.
Last June in Las Vegas, Nutanix announced all-slash systems for almost every model they have in their portfolio. Today in Vienna, the next steps were shared with the audience.
In the near future, maybe next year NVMe will be introduced in the Nutanix nodes. This will have great impact on storage performance and with a traditional hypervisor, a bottleneck for a vm can be created on storage perspective. Nutanix is redesigning AHV to prevent such a bottleneck for their hypervisor and the VMs. Performance (IO) is doubled with the new generation of AHV.
Cluster checks integrated in Prism
It was not announced as a big issue during the conference, and probably it isn’t. However, the integration of the cluster checks, probably better known as ncc, in the web console Prism, will make sure the tests are being used more often. It will be possible to schedule the tests, select particular tests, so just a subset will be executed. Results of the tests can be emailed to an administrator.
Witness for Metro Availability
A witness for Metro Availability (MA) will be introduced. For those unfamiliar with MA, it is synchronous replication leveraging a zero RPO and a near zero RTO. The witness is located on a third site and will perform an automatic site failover, when one site is not available anymore. Until now, site failover was a manual process. This automation of the site failover should lead to a even lower RTO.