Enabling an Energy- Efficient Future Internet Through Selectively Connected End Systems Jim Spadaro and Ted Brockly

Motivation Studies have found… Studies have found… 67% of office desktop computers fully powered after work hours 67% of office desktop computers fully powered after work hours Average residential computer is on 34% of the time Average residential computer is on 34% of the time Half the time no one is actively using the machine Half the time no one is actively using the machine Potential energy savings estimated $0.8 - $2.7 billion in the US per year Potential energy savings estimated $0.8 - $2.7 billion in the US per year

Motivation (cont.) Why are these machines fully powered? Why are these machines fully powered? Sporadic, occasional access: Sporadic, occasional access: User remote access User remote access Administrative access (backups, patches, etc.) Administrative access (backups, patches, etc.) Service provider access (set-top boxes, VoIP systems, etc) Service provider access (set-top boxes, VoIP systems, etc) Preservation of network state Preservation of network state

Motivation (cont.) Underlying reason: our networking principles Underlying reason: our networking principles Our architecture assumes connected hosts Our architecture assumes connected hosts Disconnectedness is dealt with as a problem Disconnectedness is dealt with as a problem

Related Work More limited solutions to power management exist More limited solutions to power management exist TCP keep-alive response proxies TCP keep-alive response proxies Dynamic Power Management and Energy Star Dynamic Power Management and Energy Star Wireless power-level tiers Wireless power-level tiers Wake-on-LAN Wake-on-LAN

Related Work (cont.) Traditional Internet Traditional Internet Assumes constant connectivity Assumes constant connectivity Lack of connectivity signals failure Lack of connectivity signals failure Delay/Disruption Tolerant Networks Delay/Disruption Tolerant Networks Emphasize connectivity in low-reliability environments Emphasize connectivity in low-reliability environments Store-and-forward architecture Store-and-forward architecture More suited to extreme environments More suited to extreme environments

Proposed Architecture Selective Connectivity Selective Connectivity Allow 3 states: Allow 3 states: On – Full connectivity On – Full connectivity Off – No connectivity Off – No connectivity Asleep – Grey area between the two Asleep – Grey area between the two Allow a host to be asleep and still have presence on the network Allow a host to be asleep and still have presence on the network Limit powering up host to “important” tasks Limit powering up host to “important” tasks

Proposed Architecture (cont.) Selective Connectivity is between the traditional Internet and DTN Selective Connectivity is between the traditional Internet and DTN Takes full advantage of reliable connectivity for high-priority tasks Takes full advantage of reliable connectivity for high-priority tasks Don’t assume that lack of connectivity implies failure Don’t assume that lack of connectivity implies failure

Chatter All incoming data is not necessarily important All incoming data is not necessarily important Computer not previously engaged on network received 6 pps over a 12-hour period Computer not previously engaged on network received 6 pps over a 12-hour period Ignore or have low-power handling of unimportant data Ignore or have low-power handling of unimportant data

Assistants Allow hosts to handle low power tasks while sleeping: Allow hosts to handle low power tasks while sleeping: Keep-alive requests Keep-alive requests Renewing DHCP leases Renewing DHCP leases Responding to ARP queries Responding to ARP queries Soft error: tell remote hosts to retry Soft error: tell remote hosts to retry High-power tasks wake host High-power tasks wake host

Assistants (cont.) Location is unimportant: Location is unimportant: Powered-on NIC Powered-on NIC Independent system Independent system Built into switches Built into switches Introduces a new point of failure Introduces a new point of failure Degree dependent on amount of responsibility Degree dependent on amount of responsibility

Exposing State Tussle between efficiency and security Tussle between efficiency and security Allows more efficient and reliable operation Allows more efficient and reliable operation Also could result in too much information being released Also could result in too much information being released

Evolving Soft State Soft state is one of the architectural successes of the Internet Soft state is one of the architectural successes of the Internet Maintaining soft state across selectively connected hosts poses a problem Maintaining soft state across selectively connected hosts poses a problem Two possible approaches: Two possible approaches: Proxyable State: maintenance of the state by assistant Proxyable State: maintenance of the state by assistant Limbo State: Recognition of distinction between “inexplicably gone” and “asleep” Limbo State: Recognition of distinction between “inexplicably gone” and “asleep”

Host-based Control How selectively connected hosts are seen by others should be a policy decision How selectively connected hosts are seen by others should be a policy decision Examples: Examples: What is exposed to which peers What is exposed to which peers What tasks are delegated What tasks are delegated What events should wake the host What events should wake the host

Application Primitives Could we design general application primitives to aid selective connectivity? Could we design general application primitives to aid selective connectivity? E.g., a generalized keep-alive that goes beyond a binary answer E.g., a generalized keep-alive that goes beyond a binary answer E.g., a way to share a list of files the host makes available on a p2p network E.g., a way to share a list of files the host makes available on a p2p network Perhaps there are not a set of primitives, but we would need to provide a program that encodes our needed functionality to an assistant Perhaps there are not a set of primitives, but we would need to provide a program that encodes our needed functionality to an assistant

Security Security issues cut across our thinking Security issues cut across our thinking Many questions: Many questions: How can tasks be securely delegated? How can tasks be securely delegated? How does a peer know an assistant has authority to act on behalf of a host or app? How does a peer know an assistant has authority to act on behalf of a host or app? How do we layer our use of cryptography to expose information needed by an assistant without exposing sensitive data How do we layer our use of cryptography to expose information needed by an assistant without exposing sensitive data

Final Thoughts Our thinking of the issues is in early stages Our thinking of the issues is in early stages We likely don’t have all models We likely don’t have all models While energy savings has been the focus, the resulting components could be useful in other contexts While energy savings has been the focus, the resulting components could be useful in other contexts E.g., mobile hosts E.g., mobile hosts