SLA Metrics and Challenges
To ensure accountability and trust within the ICN ecosystem, a set of SLA metrics and challenges are defined to measure and verify compliance:
Geolocation: The expected geographical location of the VM instance, as agreed upon in the SLA during the reservation process.
Challenge: SLA oracle nodes perform IP detection tests to ensure that Hardware Providers are not using VPNs or Proxies to mask their true geolocation.
Hardware capacity: The total amount of reserved hardware resources, including CPUs, RAM, disks and GPUs, allocated to a specific type of VM instance for a specific period of time.
Challenge: In case of a dispute regarding hardware capacity, SLA oracle nodes initiate hardware diagnostic challenges to daemons. Daemons run benchmarking tests and compare them against the machine fingerprint to verify that the actual hardware capacity is still available as specified in the SLA. The results are sent back to the SLA oracle nodes to verify compliance with the SLAs.
Hardware Availability: The percentage of time during which the hardware provider server is reachable, accessible, and functioning correctly.
Challenge: SLA oracle nodes perform periodic connectivity tests against the daemon to verify inbound/outbound traffic for specific ports and protocols as defined in the SLA.
Service Availability: The percentage of time during which the provisioned virtual machine is available, accessible, and functioning correctly.
Challenge: SLA oracle nodes send challenges to the daemon to verify that the hardware is available at the operating system level and the virtual machine is in a healthy state.
Last updated