Proposal ID: 609828
Role: Cordinator
Acronym: Orbit
Type of action: RIA
Call identifier: FP7-ICT-2013-10

Orbit: Business Continuity as a Service

Duration in months: 30
Fixed keyword 1: High Availability, Disaster Recovery
Fixed keyword 2: cloud computing
Fixed keyword 3: Infrastructure-as-a-Service
Fixed keyword 4: Fault Tolerance, Business Continuity;
Free keywords: High Availability; Disaster Recovery; Fault Tolerance; Business Continuity; Virtual Machine; Virtual Machine Monitor; Infrastructure-as-a-Service; Cloud Computing; Operating Systems;

More and more areas of public life become dependent on availability of IP based services. Banks, logistics, travel, sales and media – to name a few – are severely hit by small and specially larger and longer outages of services. As events in 2012 have shown datacenter outages can result from a number of unforeseeable events. For example a datacenter in Ireland was affected by a lightning strike. In another event a failure of data collection services had widespread effects on cloud resources in Northern-America, taking down a number of high-trafficked websites. From an EU perspective there are additional aspects to be taken into consideration. For example in case of service discontinuity a provider might replicate data to other data centers, though by doing that breaching compliance rules for data in the EU. In short: Loss of availability of mission-critical services, such as SAP HANA in-memory relational database or other data repositories with dependency on very high availability, may result in both capital and human loss. And while addressing HA for statefull and fully-consistent applications, such as SAP HANA, can be done via specialized hardware of heavy application modifications, in this project we aim at achieving this end by applying generic techniques which require neither of the two, ultimate attempting to provide HA as a standard cloud service. We farther investigate and address the full range of outage (or service downtime) causes: from single hardware failures (commonly referred to as faults) effecting a host or a network switch, up to major faults (commonly referred to as disasters) bringing down an entire data center or site.