It would be extremely helpful to show the POE status of POE switch ports.
Ditto..............
This would be great if it could maintain history on the provided power. I would love to see graphs that show "brownouts & drops" in provided power. Could be helpful in troubleshooting lockups and reboots...
great point, history would be extremely helpful, thanks!
Given all my devices I'm worried about support the standard POWER-ETHERNET-MIB (Juniper, HP) here is my ideal of support:
1) NCM gathers the state of POE and stores it in its database, so that it can be used for driving configuration change templates. e.g. if a port does not current have working POE on it then POE could be disabled. (we would use this to bring nodes into compliance with new standards for POE for example)
2) UDT gathers the same information on a per-port basis (note the POE ports do not line up with the standard IF-MIB ifIndex :-( )
I need:
a) if POE was negotiated (or if it's searching)
b) if POE is administratively disabled
c) the POE priority (high/low)
d) the POE class.
[the standard MIB only provides the negotiated class, not the actual power draw)
3) NPM on a per-device basis a multiple PSE-guage widget showing the percentage of power available and being consumed by a PSE; I can have one linear chart per PSE, but this is a pain to configure when you really need a variable number of gauges depending on how mane PSE there are in a switch/stack. a nice stacked linear gauge display showing the percent of power consumed would be good.
While I'm at it...
4) consumption of POWER-ETHERNET-MIB traps to update the power currently being consumed (the current power/PSE is reported in the trap, so there is no real need to poll for this if you can capture and process the trap data)
IME not very useful; I only need to know when a node is running out of power so we can figure out if we're going to deploy another switch (or 208V power). Once we can turn off power for things that are not supposed to be plugged in and bring the POE under change control then we should only run out of power when we fail to plan properly...
5) Historical graphs of PSE power consumption for the node
I would like to be able to generate alerts for PoE Denied.
Development info: For some reason the Juniper EXs don’t increment the denied counter OID and for some reason the HPs don’t accurately reflect detection status.
For HP e5406:
pethPsePortPowerDeniedCounter OBJECT-TYPE [1.3.6.1.2.1.105.1.1.1.12]
SYNTAX Counter32
MAX-ACCESS read-only
STATUS current
DESCRIPTION
"This counter is incremented when the PSE state diagram
enters the state POWER_DENIED." increments starting at 0, then 1, 2, 3 depending on how many times it’s unplugged and plugged in
REFERENCE
"IEEE Std 802.3af Section 30.9.1.1.8
aPSEPowerDeniedCounter"
::= { pethPsePortEntry 12 }
e.g. values:
POWER-ETHERNET-MIB::pethPsePortPowerDeniedCounter.6.142 = Counter32: 3
POWER-ETHERNET-MIB::pethPsePortPowerDeniedCounter.6.143 = Counter32: 0
POWER-ETHERNET-MIB::pethPsePortPowerDeniedCounter.6.144 = Counter32: 2
For Juniper EX:
pethPsePortDetectionStatus OBJECT-TYPE [1.3.6.1.2.1.105.1.1.1.6]
SYNTAX INTEGER {
disabled(1),
searching(2),
deliveringPower(3),
fault(4),
test(5),
otherFault(6) ”6” will tell us if a PD is being starved of power
}
Example values:
POWER-ETHERNET-MIB::pethPsePortDetectionStatus.1.32 = INTEGER: searching(2)
POWER-ETHERNET-MIB::pethPsePortDetectionStatus.1.33 = INTEGER: searching(2)
POWER-ETHERNET-MIB::pethPsePortDetectionStatus.1.34 = INTEGER: searching(2)
POWER-ETHERNET-MIB::pethPsePortDetectionStatus.1.35 = INTEGER: otherFault(6)
POWER-ETHERNET-MIB::pethPsePortDetectionStatus.1.36 = INTEGER: searching(2)
POWER-ETHERNET-MIB::pethPsePortDetectionStatus.1.37 = INTEGER: otherFault(6)
A POE budget gauge would be awesome!
We use POE for many devices, this stat would be beneficial
I've got a poller for Catalyst 4500's that may provide the budget you're looking for: Cisco Catalyst 4500 PoE pollers
been a LONG time bump :-)
monthly bump.....
keeping it going bump
Hump Day Bump
Bump!
back from vacation BUMP!!
Coming up on six years....bump?
It sure seems like simply enabling traps on the switches would provide you the POE info you seek.
I wouldn't say "no" to POE graphics on NPM Nodes, either.
Sure, a person can build a job that issues a "show power inline" command, or even a "show int count err" and learn a ton of great info, but it would be convenient to see these types of things in UDT as long as a person is looking at a device they've searched for and found in UDT.
In my environment we run POE off switchports to VoIP phones, and PC's or TC's off those phones' internal mini switch ports.
How do you envision benefits that would help you do your jobs better if you saw POE info in UDT? What would it tell you, how would it reduce MTTI or troubleshooting time? (Maybe I can benefit from your thought process and get better info . . . or maybe I can help you find an alternative that may work for you until UDT does what you need.)
traps only work when a device is added. I'm looking for information similar to vlan info on the interface. UDT simply lets me know int g2/0/22 is in vlan 230.
I would like to see a POE column that says the watts provided. This would mostly be 0 or null, and when there is a WAP or phone it could simply state the watts provided.
Bumping from the grave. The OID 1.3.6.1.2.1.105.1.1.1.6 is confirmed working on the following models:
2960X-48FPS-L
2960CX-8PC-L
3560CX-12PC-S
3560CG-8PC-S
3560-G48PS
4503-E
4506-E
The OID description fails to mention that a value of '4' means you have a PoE controller failure for that port.
