Email Subscription Notifications Suspended Temporarily
We are in the process of making navigation in the Servers and Operating Systems forums simpler and more direct. While doing this, we have to temporarily suspend email notifications for subscriptions. If you are subscribed to one or more discussion boards or blogs in the community, please check them daily to see new content. Notifications will be turned back on in a few days. We apologize for any inconvenience this may cause. Thanks, Warren_Admin
Grounded in the Cloud
cancel
Showing results for 
Search instead for 
Did you mean: 

Disaster Recovery for CloudSystem 10 Enterprise with Zerto Virtual Replication - Demo

vguddad

In the first part of this blog post we reviewed the concepts of Disaster Recovery using Zerto Virtual Replication

In this post, we will describe how HPE Helion CloudSystem Enterprise 10 (CSE 10) paired with Zerto Virtual Replication (ZVR 4.5) delivers a seamless way to protect data and applications with less data loss and downtime than with traditional snapshot backup solutions. The first combined release of CSE10 and ZVR 4.5 supports ESXi VM protection only.

Here are the steps we will follow:

  1. Disaster Recovery Setup – Review a simple DR setup
  2. Create a CSA subscription for a single VM on vCenter in Site 1
  3. Create a Zerto Virtual Protection Group (VPG) for the CSA subscription VM in Site 1
  4. Modify the contents of the VM
  5. Failover non-disruptively to Site 2 using ZVR
  6. Validate that all CS services are functional in Site 2 and the subscription VM content is recovered

Pre-requisites

  • Configure CSE 10 per the installation guide.
  • Install ZVR on the CSE 10 platform in the source and target infrastructure.

Below are the steps for installing ZVR.

  1. A Zerto Virtual Manager (ZVM) is installed at each site and interfaces with the local vCenter.
  2. The sites are paired in the ZVM
  3. The Virtual Replication Appliances (VRAs) are installed.

Disaster Recovery Setup

For ease of explanation, a simple DR setup has been chosen. There are other options possible and should be chosen based on your business and technical requirements.

We have 2 vCenter environments with identical resources but in different data centers. ZVM and VRA appliances are installed on both sites. These sites are configured as peer replication sites in ZVM.

  • Data Center 1 (Site 1) is the protected site and has the active CS 10 Enterprise installation running the control plane appliances and workloads
  • Data Center 2 (Site 2) is the recovery site and it does not have any CS 10 related virtual machines
  • DR sites can be within the same metro or region or across continents. The network latency and bandwidth between protected and recovery sites will impact your achieved Recovery Point Objective (RPO).

  • A VPG is created for the CS control plane VMs on Site 1

Create a CSA Subscription

In the Cloud Service Automation environment, we have already configured a vCenter provider for Site 1. This is where the VMs for a vCenter subscription will be created.

The VM name prefix helps identify the VM in the Zerto VPG creation process which follows next.

Create a VPG for CSA subscription

We will now create a VPG to ensure that the newly created subscription VM is protected. The direction of replication is from Site 1 (protected site) to Site 2 (failover site). The process is fairly straightforward and requires a VPG naming convention for compatibility with the CS recovery script. Either a specific ESXi host or an ESX cluster has to be selected in the failover site along with the recovery network and root folder. Reverse protection is enabled to ensure that VMs are protected even at Site 2 once Site 1 is restored to normal health. The other options are left to their default values.

The VPG created in ZVM on Site 1 is also visible in ZVM on Site 2 automatically.

Once the initial sync is complete and a recovery point objective (RPO) becomes available, the VM is under DR protection. Any subsequent changes to the VM will be replicated by Zerto to the failover Site 2 without manual intervention.

Modify Subscription VM content

To validate that modifications made post VPG creation are replicated, we will create a new text file on the subscription VM.

Failover

Now we come to the key aspect of this exercise which is triggering the failover to Site 2. Remember that the failover is not automatic, but needs to be manually re-triggered from ZVM. If Site 1 were still available, the failover could be initiated there. But in a real disaster scenario, Site 1 would be out of service. So we trigger the live failover from ZVM on Site 2 by selecting both the CS control plane VPG and the subscription VPG. Once the failover is complete, we also restart all the CloudSystem services using existing CloudSystem mechanisms.

At this point, we are able to login to the CS Ops Console, Cloud Service Automation, Marketplace portal etc. But the CSA database is not in sync with the VM details in the new vCenter environment. CloudSystem Enterprise 10.0 comes with a downloadable standalone software csrecover to integrate Cloud Service Automation (CSA) with ZVR. It is the component that updates the CSA service when the customer’s workloads fails from one compute region to another compute region. It adjusts the CSA resource provider’s access point to the VCenter in the recovery target region, and refreshes the VM information after they are recreated in the target VCenter.

Validate Recovery

We have gone through an elaborate process to setup DR, failover and run some recovery scripts. Now we need to verify that everything really works. We validate this by logging into the Marketplace portal, finding our CSA subscription, and logging in to the VM with the same credentials as earlier. We should find the text file we created earlier on that VM.

Beyond this simple active/passive DR setup one can also set up two vCenter compute regions, one on each site. The CS 10 Enterprise control plane can be installed in either of the two sites, along with the compute region, or in a third site. The federated CSE can manage and provision to both compute regions, i.e. both compute regions are active.

About the Author

vguddad

Vinod is a Senior Architect in the Cloud Technical Marketing team. He has a breadth of experience in software development, pre-sales, enablement across various technology platforms.

Events
28-30 November
Madrid, Spain
Discover 2017 Madrid
Join us for Hewlett Packard Enterprise Discover 2017 Madrid, taking place 28-30 November at the Feria de Madrid Convention Center
Read more
HPE at Worldwide IT Conferences and Events -  2017
Learn about IT conferences and events  where Hewlett Packard Enterprise has a presence
Read more
View all