If I have a small 1U or 2U server that has redundant power supplies (and let's say I get a "Y" type cord so that they actually require one socket) should I expect any charges from the data center that it will be co-located at?
A "Y" plug only provides for redundancy in the event of a power supply going out, it doesn't provide for if a breaker trips. If you want true redundancy, as has been mentioned, use two different power cords (one to each power supply) and get two different plugs on two different circuits..
With regards to redundancy, it is better to try not to create a single point of failure for your power otherwise you just defeated the purpose. Furthermore, you may want to check and be sure that both feeds are not coming from the same PDU though the same UPS, etc. Otherwise you are simple receiving two powers form the same device and still have a single point of failure issue. Also, do fall into the trap when they explain the units are "self redundant" If you take it down for maintenance or it goes off line, it is still down. Good luck.
Most centers are going to ask you to place at least two 20Amp power strips into the cabinet in order to get N+1 power. Because of the capacity fo the center they usually alternate between the A and B power grid. This means that your first and third 20amp circuit will be on grid A and and your second and fourth will be on B. The cost of having the extra grid will usually just be the cost of adding an additional 20Amps. Hope that this helps.
I disagree, not all centers drive two power strips in their cabinets. We do but honestly in some cases it would be best for the client to do the power switching further up so you dont burden the client with the cost of dual power supplies or power switching devices.
Greg Landis | Founder Jaguarpc - Keeping websites happy since 1998 Managed IT Solutions - Business hosting | Virtual Private Servers | Cloud VPS Hosting | Dedicated servers | Backup service
Follow us @ Facebook.com/Jaguarpc | Twitter: @JaguarPC | (888)-338-5261 | sales @ jaguarpc.com
Correct. Redundant power grids to redundant generators to redundant A and B internal Grids, to redundant UPS's all the way to the customers cabinet. Each progression allows for fail over cross over to the other side. Although power will fail over to the other grid, some customers want the reliability of 100% power even if the switching happens in milliseconds. For a standard server you should not have any issues with a single strip that has N+1 power upstream. But of application servers or fiber channel arrays that draw higher level's of current you may need to dual feed if you guarantee 100% to your customer.