InterCloud - Cloud based DRP
0 likes886 views
This document discusses using cloud infrastructure for disaster recovery planning (DRP). It notes that DRP is critical to prevent business downtime, loss of productivity, and other issues from outages and disasters. Implementing DRP can be expensive and complex, often requiring offsite data copies and remote resources. The document presents three common DRP strategies - cold, warm, and hot standby sites - and their characteristics. It argues that a cloud-based DRP can provide the availability of a hot standby at a lower overall cost due to pay-as-you-go pricing, easy data replication, and no hardware maintenance fees. The rest of the document provides examples of how to design and implement a cloud-based DRP
InterCloud - Cloud based DRP
- 1. INTERCLOUD CLOUD BASED DISASTER RECOVERY PLAN How to leverage on a cost effec0ve Cloud based Infrastructure to build your DRP Pierre Cérou Sales Director Pierre.cerou@intercloud.fr www.intercloud.com
- 2. WHY IS DRP CRITICAL ? Today’s organizations are exposed to a wide range of outages and disasters: fires, floods, viruses, hardware failures, corrupted data and even human error. A reliable RD strategy will prevent: • Expensive business downtime • Loss of productivity • Breach in Customer Service Agreement • Impact on reputation and/or financial assets 2
- 3. DRP DRP implementation can be expensive and complex. CHALLENGES Usually a DRP will require: • access to data copies • ressources located in remote sites. 3
- 4. DRP 3 different site strategies: CHALLENGES Cold standby Warm standby Hot standby - Involves offsite maintained - Based on disk copies of data - Sites are running systems to copies of the data on tape and driven by VM. maintain real time data and access to a similar network - Data restoration is much applications. and hardware set of faster (hours). - Hot site is available to take equipment. - Periodic testing is easier and over operations on failure of - Recovery may take hours or VM images that encapsulate the primary site. even days. Testing is complex system, apps, data and - Hot site requires upgrade and and time consuming. settings easy to synchronize maintenance like primary site between primary and standby as well as high speed network. site. - Periodic data upgrade requires fast and reliable Network. 4
- 5. DRP DESIGN OPTIONS 5
- 6. VIRTUALIZATION Virtualization doesn’t only keep hardware related costs low AND CLOUD Also helps to simplify the DRP maintenance process and provide real time flexibility. DRIVE DRP In a nutshell, main benefits are: SUCCESS • No, or very limited, CAPEX in cloud based DRP. • No long term commitment for site or equipment rental. • With cloud-based DRP, you don’t have to build a DR platform, you just rent one which will perfectly fit your requirements. • Platform is flexible. It will grow with your business for a predictable cost. • Cloud and virtualization will support multi-tier design in order to mix private and ‘virtual private’ clouds if needed. • Overall Cost Of Ownership and Operation significantly lowered by: • Pay as you go and only for what you use model • Easy data replication and no fee data import 6
- 7. Cloud based RRP will deliver the best balanced solution. DRP Cost of ‘cold standby’ for the price of ‘hot standby’ STRATEGY MATRIX High data availability Hot Hot Stand Stand by by Warm Stand by Warm Stand by Low cost High cost of ownership of ownership Cold Stand by Low data availability 7
- 8. INTERCLOU D DESIGN 9 MS EU Client 90 MS US East 1 MS SERVICE PROVIDER On-‐Net PRIVATE ZONE AS XXXX AS AAAA AS YYYY AS ZZZZ 8
- 9. DATA FLOW: INITIAL SET-UP AWS Import/export Or Avaibility zone AWS Storage Gateway EC2 Client S3 AWS Region internet 9
- 10. www.mycompany.com LIGHT PILOT: OPERATION MODE • Server Not Proxy/Caching Running Server • Pre-bundled • IMS Provisioning Application Server Critical data mirroring Light Pilot (light system core set) System Database Server Data Volume Corporate datacenter 10
- 11. www.mycompany.com LIGHT PILOT: RECOVERY MODE • Start in Proxy/Caching Minute Server • Change DNS point to AWS/ECS Servers Applica0on Server Cri0cal data mirroring Increase (light system core set) capacity if Database needed Server Data Volume Corporate datacenter down 11
- 12. ENTERPRISE 2.0 LEVERAGE ON INTERCLOUD NETWORK • Produc+vity Cloud • Low latency • Low cost per Go • Private/Secured traffic • Biz Ops & HR Cloud • High availability • Private/Secured traffic • The Virtual IT Cloud • Scalable bandwidth SINGLE LINK TO • Flexible design • Low cost per Go IMPROVED CLOUD • Private/Secured traffic EXPERIENCE • The Collabora+on Cloud • Audio / Video / Web traffic • High availability • CommiUed latency • Strong SLAs • Improved User Experience 12