cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1571
Views
1
Helpful
11
Replies

Powering a CMICR-4PT via PoE+

Joshpre
Level 1
Level 1

Hi all,

I'm currently experiencing an issue when it comes to powering a Cisco micro switch (CMICR-4PT) via PoE+ (C9300-48P). The device doesn't draw any power and never boots. when connecting via UPOE (C9300-48UXM) it powers up without any issues.

'LLDP Run' has been enabled on both the micro & main C9300-48P edge switch.

Cisco Catalyst Micro Switches Data Sheet - Cisco - Datasheet confirms powering via PoE+ is possible.

Any help is appreciated

TIA

1 Accepted Solution

Accepted Solutions

So Leo asked me to take a look at this thread.  I also have never used these micro switches.

Agreed that according to the data sheet, Table 8, it should work.  That implies that it should be able to work with 30W POE+ alone and still have 7W of passthrough POE power to spare - implying that it consumes 23W.

TAC lab repro usually takes 2 weeks (at best and just with IOL) these days - more like 4-6 weeks in most cases  so if it was quicker than that I'd be suspicious too.  Anything requiring hardware will usually be more like 6 weeks unless by some stroke of luck they happen to already have the right piece of kit there already.

So moving on to the installation guide:
https://www.cisco.com/c/en/us/td/docs/switches/lan/catalyst_microswitches/hardware/install/b-micro-hig/m-micro-product-overview.html
Table 3 there looks a bit different to the one in the data sheet!
That shows it only working with 60W or 90W PD power.  With 60W PD power it says 29W of POE passthrough power available which implies the unit is consuming 31W of the power itself which means it would not work on POE+!

Also https://www.cisco.com/c/en/us/td/docs/switches/lan/catalyst_microswitches/software/releases/15_2_8_e/configuration_guide/int_hw/b_1528e_int_and_hw_cms_cg/configuring_poe.html says "The device supports powering from uplink connected to IEEE 802.3bt Type 4." 

Table 2. PoE Budget
PoE Budget (Watts) Power Sources
55 PD (type 4)

This https://www.cisco.com/c/en/us/products/collateral/switches/catalyst-micro-switches/cat-micro-switches-wp.html#Switchpower also seems to confirm the same - it says UPOE+ powered.
So based on all that, my conclusion is that you need at least a UPOE 60W power supply for it to work.

The only other thing I can suggest to try with POE+ (on the off chance it might work) is to use CDP not LLDP (and definitely *never* have both CDP and LLDP enabled on the ports - IOS usually doesn't handle that very well).

Also make sure you're running the latest IOS just in case it's something that's been fixed/changed in a later release.

As far as TAC is concerned: if they claim to have working repro with POE+ then ask for the "show power inline", "show cdp neigh detail" and "show lldp neighbors detail" from their switch and the "show power", "show cdp neigh detail" and "show lldp neighbors detail" from their microswitch.  If you've already closed the case you can ask to re-open it within 2 weeks of closing it.

If the conclusion is that the data sheet info is wrong then submit feedback on the data sheet and they *might* correct it.  A number of the AP data sheets have wrong info on them which I've reported and they've never corrected it.  Even discussed with BU via account team and they came back with some ridiculous argument to justify having different figures on the datasheet.  They say that the data sheet reflects actual AP power consumption (which is more or less right) but the AP actually requests and reserves more power than it needs from the switch (theoretical absolute max power consumption) so the actual power budget needs to be about 30-40% higher than what the data sheet says (even though the AP never uses that power).  It's a mess and the people writing these docs have clearly never had to run a network.  The only way to confirm the POE data is to test it yourself because the data sheet info cannot actually be trusted anymore.  It wouldn't surprise me if it's the same with these - it might only use 23W but they made it reserve at least 60W anyway - you know, just in case and because Cisco says it's helping us save energy etc etc but you need to buy a UPOE switch to make it work (yes I'm getting very cynical)

View solution in original post

11 Replies 11

Leo Laohoo
Hall of Fame
Hall of Fame

Even though the Data Sheet says "DC, 803.2.bt, Cisco UPOE+®, PoE+" I believe this is not correct and the switch must be powered by uPoE (>60wac) or better.

Thanks Leo,

We currently have a TAC case open and the assigned engineer seems to think they've been able to get the micro switch working using Poe+ in a lab environment. 

Lab setup:
C9300-48P
IOSXE: 17.03.06

I have replicated this with physical kit and didn't get the same results. Has your belief come from pass experience with the micro devices? 


@Joshpre wrote:
the assigned engineer seems to think they've been able to get the micro switch working using Poe+ in a lab environment. 

TAC agent was able to get it working?  With a physical switch?  There must be some sort of mistake.  TAC Agents, from my interaction with them as early as 2013, no longer have access to physical appliances or platform. 

TAC rely heavily on IoL (IOS/IOS-XE over Linux).  This is why troubleshooting hardware issue is nearly impossible.

What "zone" is the TAC case in?  Is it in EMEA or LATAM/NAM?

Unable to confirm conclusively whether it was a physical lab or IOL, my assumption would be IOL based on how quickly they could test and have kit which replicates our live LAN. A large amount of WebEx troubleshooting was also completed with myself all failed to resolve the issue.

Would you expect to see discrepancies in functionality between physical & IOL testing?

Zone: LATAM/NAM. im based in EMEA

Was the case originally created before (or after) 13:00 UTC?

Yes

If the case was created before 13:00 UTC, then it is not EMEA nor LATAM/NAM.  

I strongly recommend to wait until 13:01 UTC and call to have the case re-queued to EMEA zone. 

Cisco customers, especially with the price/cost of TAC Support, expect good support from TAC (and not be "taken for a ride").

Thanks for the support Leo,

I'll request for our TAC to be re-logged within EMEA.


@Joshpre wrote:
I have replicated this with physical kit and didn't get the same results. Has your belief come from pass experience with the micro devices? 

I do not have any exposure to this model of switches, however, I do recall a power point presentation mentioning about uPoE to power them up because the switch (without PoE PD connected) already uses >20wac.  

So Leo asked me to take a look at this thread.  I also have never used these micro switches.

Agreed that according to the data sheet, Table 8, it should work.  That implies that it should be able to work with 30W POE+ alone and still have 7W of passthrough POE power to spare - implying that it consumes 23W.

TAC lab repro usually takes 2 weeks (at best and just with IOL) these days - more like 4-6 weeks in most cases  so if it was quicker than that I'd be suspicious too.  Anything requiring hardware will usually be more like 6 weeks unless by some stroke of luck they happen to already have the right piece of kit there already.

So moving on to the installation guide:
https://www.cisco.com/c/en/us/td/docs/switches/lan/catalyst_microswitches/hardware/install/b-micro-hig/m-micro-product-overview.html
Table 3 there looks a bit different to the one in the data sheet!
That shows it only working with 60W or 90W PD power.  With 60W PD power it says 29W of POE passthrough power available which implies the unit is consuming 31W of the power itself which means it would not work on POE+!

Also https://www.cisco.com/c/en/us/td/docs/switches/lan/catalyst_microswitches/software/releases/15_2_8_e/configuration_guide/int_hw/b_1528e_int_and_hw_cms_cg/configuring_poe.html says "The device supports powering from uplink connected to IEEE 802.3bt Type 4." 

Table 2. PoE Budget
PoE Budget (Watts) Power Sources
55 PD (type 4)

This https://www.cisco.com/c/en/us/products/collateral/switches/catalyst-micro-switches/cat-micro-switches-wp.html#Switchpower also seems to confirm the same - it says UPOE+ powered.
So based on all that, my conclusion is that you need at least a UPOE 60W power supply for it to work.

The only other thing I can suggest to try with POE+ (on the off chance it might work) is to use CDP not LLDP (and definitely *never* have both CDP and LLDP enabled on the ports - IOS usually doesn't handle that very well).

Also make sure you're running the latest IOS just in case it's something that's been fixed/changed in a later release.

As far as TAC is concerned: if they claim to have working repro with POE+ then ask for the "show power inline", "show cdp neigh detail" and "show lldp neighbors detail" from their switch and the "show power", "show cdp neigh detail" and "show lldp neighbors detail" from their microswitch.  If you've already closed the case you can ask to re-open it within 2 weeks of closing it.

If the conclusion is that the data sheet info is wrong then submit feedback on the data sheet and they *might* correct it.  A number of the AP data sheets have wrong info on them which I've reported and they've never corrected it.  Even discussed with BU via account team and they came back with some ridiculous argument to justify having different figures on the datasheet.  They say that the data sheet reflects actual AP power consumption (which is more or less right) but the AP actually requests and reserves more power than it needs from the switch (theoretical absolute max power consumption) so the actual power budget needs to be about 30-40% higher than what the data sheet says (even though the AP never uses that power).  It's a mess and the people writing these docs have clearly never had to run a network.  The only way to confirm the POE data is to test it yourself because the data sheet info cannot actually be trusted anymore.  It wouldn't surprise me if it's the same with these - it might only use 23W but they made it reserve at least 60W anyway - you know, just in case and because Cisco says it's helping us save energy etc etc but you need to buy a UPOE switch to make it work (yes I'm getting very cynical)


@Rich R wrote:
If the conclusion is that the data sheet info is wrong then submit feedback on the data sheet and they *might* correct it.  A number of the AP data sheets have wrong info on them which I've reported and they've never corrected it. 

Thanks, @Rich R

There was a time when Cisco documentation was a used as a de facto "standard". 

Sadly, `tis but a bygone era that can only be whispered in bars and gatherings.

 


@Rich R wrote:
Even discussed with BU via account team and they came back with some ridiculous argument to justify having different figures on the datasheet.

This is the latest "trend" I have observed coming out of Cisco just to escape making any improvements or doing any work.  Make some fairy tale excuse and hoping that the complaint goes away.  Developers, for instance, would respond with, "`tis not a bug because it was designed that way".

This goes back to what I've been saying, unless someone happens to be one big, fat "whale", reporting a bug or requesting a feature will always be ignored.