Inline power question for CCNP SWITCH exam study

dmease · ‎05-04-2010

Hi,

I have the Cisco Press book 'official certification guide' by David Hucaby, and have a couple of quick questions:

1) Cisco inline power (proprietary) method sends out a 340kHz test tone, and doesnt use a voltage check (as per 802.3af), as "Otherwise, other types of devices (normal PCs, for example) could be damaged". Just out of curiosity, if a switch is left as power inline auto (default), and a non-powered device is connected, does this imply that there is a flaw in 802.3af whereby the voltage check could damage the device?

2) The power inline static command is used "if you have a device that cannot interact with either of the powered device-discovery methods". Does this mean that essentially the power is left on on this port (I had previously thought that the static meant *only* that the *amount* of power was static)?

3) The book shows an example of the output from show power inline, and one of the ports is shown as Admin:static, Oper:On, Power:15.4, but this power isnt taken away from the calculated power remaining (in the book example), so this would appear that the power was 'on', but nothing connected. this would seem to answer yes to 2), but just confirming!

Any feedback appreciated!

cheers,

Giuseppe Larosa · ‎05-04-2010

Hello Dmease,

1) both detection methods (cisco and 802.3af) are not intrusive and use small signals

Detection	PSE detects if the PD has the correct signature resistance of 15 - 33 kΩ	2.7 - 10.0

see

http://en.wikipedia.org/wiki/Power_over_Ethernet

so no worries about auto mode a PC will not be damaged.

Warning: cisco documentation suggests to wait before moving a cable from a powered device to a standard NIC port to give time to the switch port to disable inline power.

This is reasonable and wise to follow this suggestion

2) I agree with you that this static option should mean unconditional power on the port, but this should be something that nowdays is not common.

clearly if the port is not connected it is an open circuit and power cannot be drawn down

there is also a question of priority, if one power supply fails the switch has to decide which ports to switch off POE on.

the static option gives priority

see

static—High priority PoE interface. The supervisor engine preallocates power to the interface, even when nothing is connected, guaranteeing that there will be power for the interface. You can specify the maximum wattage that is allowed on the interface. If you do not specify a wattage, then the switch preallocates the hardware-supported maximum value. If the switch does not have enough power for the allocation, the command will fail. The supervisor engine directs the switching module to power up the interface only if the switching module discovers the powered device.

http://www.cisco.com/en/US/docs/switches/lan/catalyst4500/12.2/46sg/configuration/guide/PoE.html#wp1071409

because the amount of power is reserved it is shown in sh power inline even if it the port is unplugged

However, I still think of this feature as uncommon and potentially dangerous for PC NICs

Hope to help

Giuseppe