Editing Greenus

=Green Venus Project= * Determine Scope * Determine Specific Objectives * Complete Design Documents ==Assumptions== The system: * must run Solaris * may run on Sun Hardware * must be connected to the Internet * must be available 24x7 * may be configured as "High Availability" (Can you tell I am a consultant Haha!!) (Can you tell I was on RFC editing teams ahahah) ==Scope Discussion== * Option 1. Basic green-er PC plugged into the grid in a standard data center environment. * Option 2. Some power derived from Solar or wind. * Option 3. Off Grid with Wireless hookup. ===Option 1 Discussion "Green-PC" === This is by far the easiest, but from a challenge/bragging rights point of view, it does not do a lot except reduce dependency on electrical requirements. Forecast Cost: $1000 ===Option 2 Discussion "Grid-Tied"=== As far as "greener" ways.. this is by far cheaper and easier than a full off-grid solution. Forecast Cost: $2500 ===Option 3 Discussion "Off Grid"=== Off-grid creates some very difficult decisions, but also provides allot of opportunity and interest. From a research point of view, compelling, but difficult and expensive. Forecast Cost: $5000 ==Evaluation Criteria== It may be interesting to develop a cost-benefit scorecard that includes the factors used to make a decision. These factors include but are not limited to: || || Green-PC || Grid-Tied || Off-Grid || ||Convenience|| || || || ||Weighted Cost|| || || || ||Grid Power Used|| || || || ||Grid Power Weighted Value|| || || || ||'''Totals''' || || || || ==High Availability== Partly as an exercise in running it, and partly to fit the 7x24 requirement, above, the system may be designed as an HA system, to allow very quick recovery from: * network / power / facility unavailability, * scheduled downtime, * severe sysamdin error, or * hardware failure. The HA design could consist of the following components: * two sites, geographically and network topologically separated * software in cluster, perhaps STONITH configuration * the abiity to turn off mirroring, patch server B, have the mirroring catch up with server A, then make Server B "active" and repeat with A.

Summary:

This change is a minor edit.

Username: