Hi vAdmins and site reliability engineers,

Welcome back!

I have already written a few blogs about Ransomware Recovery. In this post I want to shift the focus to the architecture of Site Recovery.

The implementation of this Disaster Recovery solution leverages VMware Site Recovery Manager and necessitates the presence of a minimum of two data centers. Typically, your primary data center hosts your workload while the secondary data center serves as the designated disaster recovery site.

Onboarding steps:

  1. Purchase VLR subscriptions and apply them to your organization.
  2. Download and deploy SRM and VR appliances at both sites.
  3. Pair your on-premises SRM instances to set up your DR environment. Use the cloud connection wizard in the SRM appliance configuration UI to connect your on-premises SRM deployment to the VLR cloud service. This step ensures both systems trust each other and securely share information.
  4. Activate the SRM pairs on VLR. This gives you the option to use different technology stacks for DR or RWR purposes. For existing customers, this step will replace their existing licenses with VLR subscriptions.
  5. Use the solution to protect and recover.

VMware Live Site Recovery Architecture

VMware Live Site Recovery is designed for virtual-to-virtual recovery for the VMware vSphere environment. To protect your workload, you need VMware technology in your Protected Site and your Recovery Site. This technology is based on VMware vSphere (to run your workload) but you also need additional technology that realizes Disaster Recovery. Think about securing your data (replication) and execute runbooks (orchestration).

When designing, it is crucial to consider a key requirement: ensuring each site remains independent from the other.

This independence applies to all hardware components as well as the software. That is why you will find a dedicated vCenter server within each site, but also all other components that are part of your infrastructure such as VLSR, storage, etc. etc.

This image has an empty alt attribute; its file name is image-730x259.png

this image represents all components that typically apply to this architecture.

Replication Options.

VMware Live site Recovery is not a replication solution, it monitors and interacts with replication solutions.

There following options are supported:

1: vSphere Replication/Host based replication

2: Array Based Replication*

3: vVols Replication

*Array Based Replication must be on the compatibility list and support both replication 

Virtual Volumes:

In case you are not familiar with vSphere Virtual Volumes, see this solution overview.

vVols uniquely shares a common storage operational model with vSAN, the market leading hyperconverged infrastructure (HCI) solution. Both solutions use storage policy-based management (SPBM) to eliminate storage provisioning, and use descriptive policies at the VM or VMDK level that can be applied or changed in minutes. SPBM accelerates storage operations and reduces the need for specialized skills for storage infrastructure.

Source: VMware.com

vVols provide many of the benefits of vSphere Replication (like granularity of protection, per-VM replication)with the benefits of array-based replication including multi-VM consistency groups, storage policy-based management, etc).

Replication based on vVols use the VASA provider instead of the Storage Replication Adapter (SRA) . Did you know that vVols enable automatic VM protection, just by associating a VM (that is already on a vVol) with a policy that includes replication, or by placing a VM onto a vVol and associating it with a replication policy, the VM will be automatically protected

Array-based replication

When you replicate data (from the protected site to the Recovery site) based by array-based replication, you need to have similar hardware on both sites and the replication is offered by the storage array. The Storage array will utilizes the Storage Replication Adapter (SRA) for communication between Site Recovery Manager and the storage array.

vSphere replication

VMware vSphere Replication is a hypervisor-based, asynchronous replication solution for vSphere virtual machines. It is fully integrated with VMware vCenter Server and the vSphere Web Client. vSphere Replication delivers flexible, reliable and cost-efficient replication to enable data protection and disaster recovery for all virtual machines in your environment.

Source: VMware.com

vSphere Replication Details

Compression

When setting up replication for one or multiple VMs, you have the option to activate end-to-end network compression, which is initially turned off. Enabling this functionality is straightforward – a quick checkmark is all it takes, with no intricate settings to adjust. vSphere Replication employs FastLZ for compression, offering quicker speeds and lower CPU usage compared to similar algorithms while maintaining comparable compression rates ranging typically from 1.6:1 to 1.8:1. The data replicated undergoes compression on the primary side and remains compressed until it reaches the target storage. Although it reduces network bandwidth consumption, it does necessitate extra CPU resources on both the source and target ends. This added CPU demand is typically manageable since most vSphere setups have ample CPU capabilities to accommodate the increased workload, with the extent of CPU utilization depending on the number of VMs and volume of data being replicated.

Encryption

Network encryption is available for both new and ongoing replications to bolster the security of data transmission. By enabling encryption on replication traffic, you can ensure a secure flow from the source ESXi host to the target site’s datastore. The vSphere Replication appliance seamlessly deploys an encryption agent on the source ESXi hosts, utilizing the robust TLSv1.2 protocol for network encryption.

MPIT Recovery

When it comes to restoring virtual machines to specific points in time (PIT), like the most recent consistent state, Virtual Recovery (VR) offers a unique feature. The system enables the retention of up to 24 points-in-time, allowing for a rollback to any of these points post a failover incident. These points are saved as snapshots on the VM located at the TARGET site. Upon VM recovery, these snapshots are available in its snapshot tree.

A note to remember is that VMware vSphere Replication is not the only solution for VMware Site Recovery. VMware Site Recovery incorporates replication utilizing this method. Through the integration of this replication technology along with various components such as a runbook, Site Recovery delivers a comprehensive disaster recovery solution.

Quiescing

VR supports application quiescing for applications that require it. For the majority of contemporary applications, crash consistency suffices. We extend our assistance to applications that demand it, although its availability is confined by the OS and vmTools. Refer to the compatibility guide for more information.

After the initial replication, only changes to the VM are replicated to minimize network bandwidth consumption.

Did you know that with Site Recovery you can achieve a RPO as low as 1 minute per VM?

We are approaching the end of this post. A new post is currently being written with a focus on defining protection groups, recovery plans and automatic VM protection so stay tuned!

End of this post.

Disclaimer: Please note that the views expressed in this blog are solely my own and should be treated as personal opinions. This content does not hold any legal or authoritative standing.

Leave a Reply

Your email address will not be published. Required fields are marked *