During last VMUG event, customers asked few interesting questions around VVol’s support for other important HDS and VMware technologies that are important for high availability, stability and simplified management. HDS announced the release of advanced Virtual Volumes solutions that introduce significant enhancements to enable customers accelerate the adoption of this transformational storage technology. VVol framework has been a work-in-progress in VMware and partners’ engineering labs for many release cycles now, obviously so, as any such technology that fundamentally changes the way storage services are consumed and managed, would take time to become mainstream. IT analyst Evaluator Group has released a white paper that beautifully illustrates VVol concept here. Since VMware and HDS launched VVol in early 2015, there have been few sticking points for wide adoption, especially array-based replication, availability of VASA provider, DR automation and testability, vROPs support. With the release of VASA provider 3.3, HDS has taken a giant stride in delivering on customer expectations in their VVol journey. In this post, I would cover few notable additions to our solution, which have been repeatedly requested by customers before they aggressively leverage the operational simplicity of the VVol model.
1. High availability of VASA Provider :
What would happen to VVol-VMs, if VP goes down for some reason? Will I be able to spin new VMs and deliver storage services, while my VP is down? These have been few common refrains from customers. The simple answer is your VMs will continue to run uninterrupted, as IO flows through in-band path via protocol endpoint. However, unavailability of VP will prevent you from provisioning new VMs or performing any data operations, as they are enabled by VP. To maximize the VP uptime, HDS now allows customers to configure vSphere Fault Tolerance to ensure continuous availability of VP. Please remember that for FT to work, you would require to provision two hosts in active-active cluster mode and two VMs will work in lockstep. All the configurations and changes on primary VMs will be mirrored on the other VM. As shown in figure 01, If primary VM goes down, the VM on other host will take over and the VVol operations will continue with zero (minimal) impact. If your shop has higher tolerance for downtime, you have less resource-intensive vSphere High Availability (HA) option.
2. Array-based replication :
To deal with unanticipated business disruptions including array breakdown, site-failure and ensure disaster recovery with minimum RPO, customers require array-enable replication functionality. Without business continuity and disaster recovery plans, customers run the risk of loss of business, compliance violation, lawsuits or even going out of business. One of the frequent questions that we hear is how can remote backup and recovery be accomplished with VVols. You will now have prescriptive guidance and toolset to configure data volume replication between two sites using VVol datastores by leveraging Hitachi Universal Replication or Hitachi File Replication for NFS VVol. Image 3 shows the widget for setting up cross-site VM-data transfer for fail-over. You can fail-back VMs to the protected site at a later point of time, once it is restored.
3. Single Sign-on :
VASA provider 3.3 introduces Single sign-on feature that acts as authorization broker for simplified access to VVol environments. Now, you would have anytime, anywhere access to securely log into your VVol set-up across block and file VASA providers, with a single set of credentials.
4. Simplified VASA Provider Installation :
We are introducing integrated HCS (Storage management software) for VVol within the virtual VP appliance to further simplify the VVol deployment and operations. Now, you need to install just one VM, which includes the functionality of HCS-for-VVol and VASA provider. Hitachi VASA provider simplifies deployments and management of vSphere Virtual Volumes on different storage systems by providing unified installation package for block and file systems.
If you have questions related to architecture, scale, compatibility matrix, deployment, a detailed FAQ can be found . If VMworld figures in your itinerary in next few months, do register for breakout session STO 8840 - Deploy Scalable Private Cloud with vSphere Virtual Volumes