Depending on the the nature of the modeled equipment, redundancy settings, and power path setup in IT Advisor, the capacity calculations can get quite complex. If you have a similar setup as one of these simplified examples, they may help you understand where the numbers in the tooltips come from.

[EcoStruxureIT]

Depending on the the nature of the modeled equipment, redundancy settings, and power path setup in IT Advisor, the capacity calculations can get quite complex. If you have a similar setup as one of these simplified examples, they may help you understand where the numbers in the tooltips come from.

Example 1: Multiple power supplies

The data center solution includes:

• 1 large server, 3000 W load
• 6 power supplies

Capacity values:

In previous versions of ITA, the Estimated Load would be very high. In the current version, it now works as follows:

• Estimated Load: 500W on each connection (even load)
• Estimated Load: 1000W on each connection (uneven load)
• Power Supply Unit Uncertainty: 500W  (uneven load)
  
  If the server only uses half of the PSUs, the system must reserve up to: (3000W) / (6/2) = 1000W on each connection.
  This implies a Power Supply Uncertainty of 500W (because 1000W - Estimated Load = 500W).

To minimize the impact of PSU Uncertainty on the PDU, the PSU Uncertainty be only be carried upstream to the PDU in case of phase shifts.

Blade solutions:

If a blade enclosure includes bays powered by specific PSUs, the enclosure may only be using e.g. 2 out of 6 PSUs because the other bays are empty. In this case, the PSU Uncertainty reservation done by the system might be too small!

Example 2: Classic 2N and multiple power supplies

The data center solution includes:

• 1 large server, 3000 W load
• 6 power supplies
• 3 connections to each of two separate rack PDUs

Capacity values:

• Estimated load: 500 W on each connection
• Failover load: 500 W on each connection

If the failover load can carry the full power supply uncertainty (as in this example), power supply uncertainty is not added to the calculations.

Example 3: Reservation for Distribution Redundancy

The data center solution includes:

• 1 server, 1000 W load
• 2 connections to two different PDUs (L1 and L2)
• Both PDUs are connected to the same parent PDU

Capacity values:

Load overview
Load type and phase	PDU 1	PDU 2	Parent PDU
Estimated Load, L1	500 W	0 W	500 W
Estimated Load, L2	0 W	500 W	500 W
Failover Load, L1	500 W	0 W	-
Failover Load, L2	0 W	500 W	-
Reserved for Distribution Redundancy, L1	-	-	500 W
Reserved for Distribution Redundancy, L2	-	-	500 W

On the parent PDU there is no failover load because additional load will not move to this PDU in case of a downstream failure. However, a phase shift might happen in case of a downstream failure because of the mixed phases on the server. (If the server would have been connected to only one single phase, there would be no need for reserving extra capacity.)

Example 4: Reservation for Distribution Redundancy with opposite effects

The data center solution includes:

• 1 server, 1000 W load
• 2 connections to two different PDUs (L1 and L2)

• 1 server, 600 W load 
• 2 connections to two different PDUs (L2 and L1)

Capacity values:

Load overview
Load type and phase	PDU 1	PDU 2	Parent PDU
Estimated Load, L1	500 W	300 W	800 W
Estimated Load, L2	300 W	500 W	800 W
Failover Load, L1	500 W	300 W	-
Failover Load, L2	300 W	500 W	-
Reserved for Distribution Redundancy, L1	-	-	200 W
Reserved for Distribution Redundancy, L2	-	-	200 W

The two servers partly cancel out the need to reserve extra capacity.