VMworld 2015: Disaster Recovery of Tier 1 Apps using SRM & vSphere 6

This session was decent and few people can complain about its lack of technical detail. Rather, the struggle here was against losing the forest for the trees. Perhaps a winning strategy could have been starting with the demo at the end and using it as the use case from which to explain each component and contrast old methods vs. new features.

The latest versions of SRM 6.1, vSphere 6, and NSX 6.2 (as well as SRM Air) bring incredible new capabilities to disaster recovery and orchestrated failover plans. It’s definitely at a point worth engaging a trusted VMware partner who can understand your specific environment and then architect one or more solution sets using the many VMware options, as well as 3rd-party products.

Live Notes & Commentary

State of the Union: Only 50% of organizations have a formal DR plan

High Availability & Disaster Recovery for Business Critical Applications

  • Site Application Availability: SRM, DR to Cloud using SRM
  • Local Application Availability: vMotion, svMotion, FT, HA
  • Data Protection: vSphere Replication, vSphere Data Protection, Storage APIs – Data Protection

SRM Overview & What’s New

  • Orchestration & automation solution for DR
  • SRM is *not* a replication solution—it works with array or software-based replication

Async Replication Based DR Topologies

  • Active-Passive Failover: Dedicated resources for recovery
  • Active-Active Failover: Leverage recovery infrastructure for test/dev
  • Bi-directional Failover: Production at both sites (only 50% to recover)
  • Shared Recovery Sites: Many-to-one, used for ROBO

New in SRM 6.1: Storage Profiles/Policies

  • Backed by one or more datastores: tags or storage capabilities
  • VM associated with a storage profile
  • Compliance check
  • VM can be associated with SP either through Storage migration (svMotion) or on creation

VMware Metro Stretch Clusters

  • Applications protected against localized site failures
  • Proactive live migrations of apps for continuous availability
  • Automated recovery of workloads

Challenges of Stretched Clusters

  • Availability of a single vCenter managing both sites
  • DRS and HA are not site aware
  • No orchestration or testability

In SRM 6.1: Integration of Stretched Storage & Orchestrated vMotion

  • Best of both stretched storage & SRM
  • Unified plan for disaster avoidance, DR, and mobility

Manual mapping of network resources (NSX)

  • In NSX 6.2, supports syncing rules across sites
  • Universal logical switch makes network mapping unnecessary
  • Enables cross-vCenter migration and recovery

Examples:

  • Oracle Data Guard
  • SQL AlwaysOn
  • Exchange DAG

VMworld 2015 | Monday | DR of Tier 1 using SRM and vSphere 6 (STO4525)
Kannan Mani, GS Khalsa

Be First to Comment

Leave a Reply