Around the Storage Block
StorageExperts

What is the best way to deliver VMware availability? Think HPE Nimble Storage dHCI

HPE offers a dHCI environment capable of providing high availability for the environment through synchronous replication, and automation of the failover process through the implementation of Peer Persistence. With HPE Nimble dHCI, we can achieve high availability of the hypervisor management layer through the vCenter HA – exactly the kind of experience our customers are looking for in a hyperconverged solution.

HPE-Nimble-Storage-dHCI_replication_vCenter-HA_blog_small_GettyImages-1263041106.jpgWhenever I talk to our customers who are looking for an HCI solution, I hear several reasons as to why they are doing so – such as facilitating infrastructure management, speeding up the deployment of virtual workloads, or improving the control and efficiency of their VM environment. But in the end, their primary reason for seeking a hyperconverged solution is all about their customer experience.  

HPE Nimble Storage dHCI enables customers to implement a synchronous replication environment capable of performing automatic switchover (ASO). They can also rely on the high availability of vCenter[1] through the VMware vCenter high availability (HA) feature.

The elements of the HPE Nimble Storage dHCI solution

VMware vCenter HA

VMware vCenter HA was introduced in vSphere 6.5, and provides robust general-purpose protection against hardware and operating system failures for vCenter Server instances running on a virtual machine. Once configured, vSphere HA monitors hosts and virtual machines, and takes the user-configured action with or without vCenter Server availability.

A vCenter HA cluster consists of three vCenter Server instances. The first instance, initially used as the active node, is cloned twice to a passive node, and once to a witness node. Together, the three nodes provide an active-passive failover solution.

All three nodes are necessary for this feature to function. The node responsibilities are compared below.

Active node

  • Runs the active vCenter server instance
  • Uses a public IP address for the management interface
  • Uses the vCenter HA Network for replication of data to passive node
  • Uses the vCenter HA network to communicate with the witness node

Passive node

  • Is initially a clone of the active node
  • Constantly receives updates from – and synchronizes state with – the active node, over the vCenter HA network
  • Automatically takes over the role of the active node if a failure occurs

Witness

  • Is a lightweight clone of the active node
  • Provides a quorum to protect against split-brain situations

You can find more information on VMware vCenter and vSphere here.

Peer Persistence and replication for HPE Nimble Storage dHCI

As we know, HPE Nimble Storage is the storage foundation of HPE Nimble Storage dHCI.

In November of 2018, Hewlett Packard Enterprise (HPE) announced synchronous replication with automatic failover for HPE Nimble Storage, keeping the widely recognized name of Peer Persistence. The HPE Nimble Storage Peer Persistence feature was developed for applications that require zero to near-zero recovery point objectives (RPO) and recovery time objectives (RTO). By using multi-site synchronous replication with automatic switchover, Peer Persistence enables business continuity by keeping your applications online all of the time.

IP replication network.png

 Please note that HPE Nimble Storage arrays have two controllers, one active and one standby. When an initiator writes data to a volume on the array, the data lands in non-volatile random access memory (NVRAM) on the active controller. It is then immediately replicated across a chassis backplane to the NVRAM in the standby controller.

The array waits until the data is successfully mirrored to the NVRAM in the standby controller before it acknowledges (to the initiator) that the data has been written. For synchronous replication, HPE Nimble Storage arrays extend the number of copies of data that are written before the write operation is acknowledged. If an initiator writes data to a volume that is configured for synchronous replication, the data lands in the NVRAM of the active controller on the array that is presenting the volume to the initiator.

Next, the data is mirrored across the chassis backplane to the standby controller in the same array. Then it is copied across an IP address interface to the active controller in the second HPE Nimble Storage array. Finally, the data is mirrored across the second array’s backplane into the NVRAM on the second array’s standby controller. Only after the data is successfully copied to all of these locations does the first array acknowledge to the initiator that the data has been written.

The HPE Nimble Storage automatic switchover feature can be used with synchronous replication to enable fully automatic failure recovery. When ASO is enabled and a switchover occurs, the destination array that serves as the mirror for a volume collection is switched automatically, and application I/O is redirected to the partner without host-side reconfiguration.

Conclusion

With the combination of all these elements, we have a dHCI environment capable of providing high availability for the environment through synchronous replication, automation of the failover process through the implementation of Peer Persistence. Finally, we can achieve high availability of the hypervisor management layer through the vCenter HA – exactly the kind of experience our customers are looking for in a hyperconverged solution.

Thanks for reading my blog!  I look forward to your questions and comments on the content.

Linkedin photo.jpgMeet new HPE Blogger Deivis Augusto Carobin. Deivis is an enthusiastic HPE Storage Ambassador and HPE Storage Solutions Architect & HPE Master ASE – an Advanced Server Solutions Architect and HPE Nimble Storage Instructor.

For over twenty years he has been developing projects involving high availability and performance solutions featuring a variety of HPE Storage technologies. He currently lives and works in the southern region of Brazil.

You can connect with Deivis on LinkedIn!


Storage Experts
Hewlett Packard Enterprise

twitter.com/HPE_Storage
linkedin.com/showcase/hpestorage/
hpe.com/storage

 

[1] HPE Nimble Storage dHCI is managed by integration with vCenter.

About the Author

StorageExperts

Our team of Hewlett Packard Enterprise storage experts helps you to dive deep into relevant infrastructure topics.