Cloud ready discussion
Download as PPTX, PDF0 likes380 views
The document discusses design considerations for building cloud-ready infrastructure that is scalable, resilient with no single points of failure, and able to leverage multiple datacenter architectures. It provides examples of components that could be used, including storage solutions from EMC that leverage FAST caching and tiering, Cisco UCS compute blades, and replication technologies like RecoverPoint and VPLEX to provide disaster recovery and continuous availability between datacenters. Management and monitoring tools from VMware like vCenter Operations and Capacity IQ are also discussed.
Cloud ready discussion
- 1. Cloud Ready: an example based discussion Neil S. Stein January 2013
- 2. Design Considerations Infrastructure capacity needs to be easily scalable both vertically and horizontally Management overhead should not scale with it Very resilient with no single points of failure Multiple site architectures leverage best-of-breed virtualization and replication technologies Extendable from private to: public and hybrid cloud models (federated) Components sized for expansion and growth 2
- 3. Design Considerations: Hot / Warm Provide multi datacenters with capability for failover in either direction Run production at each site or use secondary for test/dev Easily perform DR testing and failover Lower cost while still providing excellent resiliency Components Storage – EMC Storage, Data Domain /Avamar Replication – RecoverPoint/SE Compute – Cisco UCS Management – vCenter Operations, CapIQ, Chargeback
- 4. Hot / Warm (discussion example only)
- 5. Design Considerations: Continuous Availability Provides multiple active datacenters with full production capabilities and HA resiliency Continuous Availability more than Disaster Recovery Allows for horizontal / vertical scaling Components Storage – EMC Storage, Data Domain/Avamar Replication – VPLEX Compute – Cisco UCS Management – vCenter Operations, CapIQ, Chargeback
- 6. Continuous Availability (discussion example only)
- 7. Storage (discussion example only) VNX 5500 or 5700 Maximum Cap: 250 or 500 Drives Proposed: 200 GB EFDs for FAST Cache or Tier 0 300 GB & 600 GB 15k SAS for Tier 1 2TB NL-SAS for Tier 2/Archive Connectivity = Any protocol desired (FC, FCoE, iSCSI, NFS, CIFS) Software licensing FAST VP FAST Cache Local and Remote Protection Application Protection
- 8. Compute (discussion example only) Cisco UCS Maximum Cap: ~160 Blades (Current) Proposed: ~24 Blades Future State Each blade is 12-cores and 96GB of RAM Benefits = Density, Service Profiles, Management. Connectivity Network – 10Gb to Network Core Storage – 8Gb FC to FC Core All Compute running vSphere 5
- 9. Replication RecoverPoint (discussion example only) RecoverPoint provides replication and journal rollback capability far exceeding other replication technologies Proposal for Hot/Warm – RecoverPoint CRR (Continuous Remote Replication) at both sites Implementation will include 2-4 appliances at each site utilizing the simple array splitter technology No invasive FC fabric configuration needed Also proposed are application plugins allowing for application consistent rollback, if desired. Integration with Site Recovery Manager
- 10. Replication – VPLEX (discussion example only) VPLEX provides for active/active storage Combine that with vSphere or other cluster-aware applications and you have a very resilient environment Allows for easy horizontal scaling of arrays once you reach capacity Future solution is single-engine (dual storage processor) VPLEX Geo at each location
- 11. Backup/Recovery Data Domain/Avamar Multiple central tier Remote capabilities
- 12. Additional Tools From VMware vCenter Operations – Provides sophisticated, but very easy to use, virtual environment performance monitoring and prediction vCenter Capacity IQ – Allows for trending and “what if” analysis vCenter Chargeback – or “showback” vCloud Director – enable cloud self-service from internal customers. Multi-level monitoring, alerting capabilities 12
- 13. Team Knowledge Technology areas: VMware, EMC, Cisco UCS Implementation in Healthcare Compliance focused
- 14. Ongoing Services Portfolio Health Checks Trend Analysis Capacity/Utilization Reporting Continuous Monitoring Proactive Maintenance Engineering and Administration Day to day management Operational tasks Patching, code revisions Reactive Support On demand support Ticketing Escalations Severity Levels Root Cause Analysis On-going Architecture and Advisory Services Continued consulting, design and architecture services Implementing recommended changes
- 15. Open Discussion