Infinidat Blog

Introducing Active-Active Replication with VMware vSphere Metro Storage Cluster for the Infinidat InfiniBox

High Availability

Infinidat InfiniBox Active-Active replication is now listed on VMware Partner Verified and Supported Products (PVSP) as a verified solution with vSphere Metro Storage Cluster (vMSC).

The vMSC solution, combining storage array-based synchronous replication with VMware vSphere capabilities, enables vSphere clusters to spread across geographical locations (aka "Stretched Clusters"). Deploying vMSC together with InfiniBox Active-Active replication provides the most resilient solution for protecting applications data, availability and enables an extremely fast recovery during a failure scenario, even in the event of a full site failure.

vMSC solutions are classified into two distinct types based on the fundamental difference in the way the vSphere hosts access the storage array, which influences design considerations.

  • Uniform access - vSphere hosts on both sites are all connected to the storage systems across both sites. Datastore paths are presented to each vSphere host from the InfiniBox systems on both sites.
  • Non-uniform access - vSphere hosts at each site are connected only to the local storage system at the same site. Datastore paths are presented to each vSphere hosts only from the InfiniBox system on the same (local) site.

InfiniBox supports both uniform and non-uniform host access types (host typology), allowing customers to deploy vMSC with the design that fits best to their environment and needs.

Maximizing Applications' Resiliency with InfiniBox Active-Active Replication and vMSC

InfiniBox Active-Active replication, providing zero-RPO and zero-RTO, was introduced in R5 to maximize mission-critical business-services availability and to enable applications to keep operating even through a complete site failure:

  • Symmetric synchronous replication solution, allowing applications to be geographically spread across data centers.
  • Fully integrated into InfiniBox, allows simple and effortless management.

In addition to the high availability capabilities, using vMSC with InfiniBox Active-Active replication enables fully active and workload-balanced data centers.

High Availability, Minimal Performance Impact.

As vMSC leverages InfiniBox innovative and high-performance replication mechanisms, the replication RTT-added latency is minimized, allowing achievement of a high level of resiliency and availability, with a minimal performance impact.

InfiniBox can also intelligently hint to the vSphere hosts about which are the optimal paths to serve I/O. This is crucial when configuring vMSC with a Uniform storage access, as the hosts are presented with typically less-optimal paths to the datastore from the remote InfiniBox system.

Storage Failed, All Applications Are Still Running

Using InfiniBox and vMSC, customers gain a transparent and automatic failover between data centers in case of failures. Deploying vMSC using the uniform storage access provides the most resilient solution, allowing all vSphere hosts to continue to access the storage (datastores) even in the case of an entire InfiniBox system failure.

In cases where vSphere hosts in a site encounter a storage failure, they will automatically shift I/O to the non-optimized paths to the remote InfiniBox system in the other site, resulting in a completely transparent and uninterrupted fail-over. 

  • In the following example System A completely fails (becomes "offline"), while the vSphere hosts in Site A can still access the storage through the non-optimized paths to the InfiniBox system in Site B.
  • All VMs running in Site A continue to run uninterrupted.

When configuring vMSC with non-uniform storage access, the vSphere hosts in each site can access the datastores only through the local InfiniBox system - the system that exists in the same site. Therefore, if a system in a specific site fails, the vSphere hosts in that site cannot continue to access the datastores. Yet, the affected VMs will be quickly and automatically recovered to the vSphere hosts on the other site by vSphere HA.

InfiniBox Replication - Automatic Failover and Recovery

InfiniBox Active-Active replication failover and recovery is completely automatic - no intervention is necessary to trigger a re-sync to recover replication. If an InfiniBox system was disconnected, the replication will internally fallback to async mode. Once the connectivity between the systems is restored, synchronization jobs will start replicating the missing data to the lagging system.

As the volumes are nearly in sync, they will smoothly transition to Sync replication mode, with no I/O disruption. The host paths to the lagging side will be automatically restored, allowing the hosts to perform I/O operations through both systems again.

Simplifying Active-Active Replication Management for VMware Administrators

InfiniBox Active-Active replication management is even easier when using the INFINDAT Host PowerTools plugin for VMware vCetner.
Host PowerTools for VMware maximizes ease-of-use for the VMware administrators by providing storage and host automation, best practice validation and comprehensive end-to-end visibility into InfiniBox, all consolidated directly within vCenter.

The following image illustrates the presentation of a datastore residing on an InfiniBox volume undergoing Active-Active replication.

  • Complete end-to-end visibility from the datastore level → both InfiniBox system → corresponding volume, simplifying the tracking of the participant objects and components.  
  • Information about each system ALUA state allowing to easily understand which InfiniBox is serving the I/O.
  • Configuring the desired ALUA state for each InfiniBox system is also available directly within the screen.

For more information on Infinidat Host PowerTools for VMware go to Infinidat Host PowerTools. To learn more about the InfiniBox integration with VMware environments, download “An Introduction to Infinidat InfiniBox and VMware Integration” from our Resource Library.

About Ido Dubovi

Ido Dubovi is the Director of Product Manager and VMware expert responsible for driving Infinidat’s partner ecosystem integration. His work covers proprietary and open source platforms. An observant market watcher and analyst, he is a key influencer of INFINIDAT’s solutions roadmap. His prior roles include consultant for numerous IT verticals and storage architect for datacenter and cloud deployments.