Header Only - DO NOT REMOVE - Extreme Networks

AP7522 POE issue


Please help, today i have being setting up and configuring 10 AP7522s i have got 7 all working correctly but for some reason 3 have POE issues they are constantly rebooting and do not say up. I have changed the cable, directly connected the AP tried different POE config on the switchports but nothing.

The switch is providing POE so all i get is the below:

*Oct 16 10:56:16.747: %ILPOWER-7-DETECT: Interface Fa2/0/23: Power Device detected: IEEE PD
*Oct 16 10:56:17.787: %ILPOWER-5-POWER_GRANTED: Interface Fa2/0/23: Power granted
*Oct 16 10:56:22.435: %LINK-3-UPDOWN: Interface FastEthernet2/0/23, changed state to up
*Oct 16 10:56:24.463: %LINEPROTO-5-UPDOWN: Line protocol on Interface FastEthernet2/0/23, changed state to up
*Oct 16 10:56:40.723: %LINEPROTO-5-UPDOWN: Line protocol on Interface FastEthernet2/0/23, changed state to down
*Oct 16 10:56:45.744: %LINEPROTO-5-UPDOWN: Line protocol on Interface FastEthernet2/0/23, changed state to up
*Oct 16 10:57:01.191: %LINEPROTO-5-UPDOWN: Line protocol on Interface FastEthernet2/0/23, changed state to down
*Oct 16 10:57:05.926: %LINEPROTO-5-UPDOWN: Line protocol on Interface FastEthernet2/0/23, changed state to up
*Oct 16 10:58:03.314: %LINEPROTO-5-UPDOWN: Line protocol on Interface FastEthernet2/0/23, changed state to down
*Oct 16 10:58:04.328: %LINK-3-UPDOWN: Interface FastEthernet2/0/23, changed state to down
*Oct 16 10:58:03.390: %ILPOWER-7-DETECT: Interface Fa2/0/23: Power Device detected: IEEE PD
*Oct 16 10:58:04.396: %ILPOWER-5-POWER_GRANTED: Interface Fa2/0/23: Power granted
*Oct 16 10:58:09.030: %LINK-3-UPDOWN: Interface FastEthernet2/0/23, changed state to up

it just repeats over and over

Please help thanks

38 replies

Thanks, I already checked this and I couldn't see anything I have double checked again still nothing I tried to ping the MINT ID but still no response.

I will proceed with getting the AP replaced.

Thank you for all your help!
Userlevel 3
One last thing that I'd try, assuming there are other APs at that site that are functioning.:

  1. connecting to a working remote AP at the site via ssh/telnet.
  2. use the command "show mint neighbors" - You will see all APs on the same Layer2 and maybe some Layer 3 neighbors depending on how your config is setup. The Mint-ID which is XX.XX.XX.XX contains part of the MAC address of the AP (the last 3 digits); eg MAC address 5C-0E-8B-11-22-33 will show up as Mint ID: 0B.11.22.33. If it doesn't show up, then the AP is probably hosed.
  3. Try pinging the mint address using the command: "mint ping XX.XX.XX.XX" if you get a response continue to next step.
  4. Now the cool part: use the command: "connect mint-id XX.XX.XX.XX" - This will open a Layer 2 CLI session with the target AP. This is especially useful to get into an AP that doesn't get an IP address for some reason.
  5. You can now troubleshoot / factory default the problem AP.
A couple comments on the config:

  • It appears that the switch port vlan 4 config is consistent, the problem lies elsewhere.
  • Why is there an IP address listed in interface vlan 4? If you are using dhcp, it should read "ip address dhcp"
  • The DHCP helper is usually configured on the switch, not on the AP. Only the switch has the switch virtual interface (SVI) for vlan 4 should have this configured.
  • "allowing" all VLANs into the AP can cause issues if there are APs on the other VLANs that could be visible over layer 2.

  • The APs IP addresses are Static they will be assigned the IP listed under that VLAN
  • DHCP helper address is configured on the SVI of the switch as well as in the profile just for peace of mind at the moment.
I believe the AP is faulty I have a total of 200 APs on multiple sites all with similar if not the same configuration, they all function correctly. The 10 on this site are all up and functioning correctly with the same configuration listed above, but this one just doesn't obtain an IP address ever.
Userlevel 3
A couple comments on the config:

  • It appears that the switch port vlan 4 config is consistent, the problem lies elsewhere.
  • Why is there an IP address listed in interface vlan 4? If you are using dhcp, it should read "ip address dhcp"
  • The DHCP helper is usually configured on the switch, not on the AP. Only the switch has the switch virtual interface (SVI) for vlan 4 should have this configured.
  • "allowing" all VLANs into the AP can cause issues if there are APs on the other VLANs that could be visible over layer 2.
AP interface config:

interface ge1
switchport mode trunk
switchport trunk native vlan 4
no switchport trunk native tagged
switchport trunk allowed vlan 1-4094
interface vlan4
description
ip address XX.XXX.XX.XX/XX
ip dhcp client request options all
ip helper-address XX.XXX.XX.XX

Switchport config:

interface FastEthernet2/0/23
description Link to APXX
switchport trunk encapsulation dot1q
switchport trunk native vlan 4
switchport mode trunk
no lldp med-tlv-select power-management
end

Thanks
Userlevel 3
The fact that you are seeing LLDP info on the switch port indicates that a) the AP booted up OK; b) there is nothing physically wrong with the interface. Both of these facts point to a configuration mismatch of some sort.

I've seen many instances where the AP has been configured incorrectly either by accident (typos) or due to unintentional behavior caused by profile overrides. The AP's configuration ends up such that the network interface does not match the switch port setting.

Please post the output of the AP's exected interface config. You can do this from the controller CLI with "show run device device_mac_address" and include all the lines for "interface ge1, interface vlan, etc".

Also include the switch's interface config for that interface.
DHCP is Vlan 4 and switchport is trunking Vlan 4, plenty of DHCP leases available within the scope, do we agree this AP is faulty?

Thanks for everyone's help so far, it's been really helpful!
Userlevel 4
Native vlan is 4 and DHCP is vlan 1? The AP need a untag VLAN to get a IP address, if it's a clean AP. Do you have free DHCP IPs?

Check for the right VLAN as untag and available DHCP address. If this are booth right, the AP is damage.
I have 10 of these access points 3 on one vlan and 7 on another all of this aps are configured the same and 9 of the 10 have been adopted and are functioning correctly​ The switchport config is Switchport mode trunk Switchport trunk native vlan 4 This access point isn't getting a lease from DHCP no matter which vlan I put it in. Thanks
Userlevel 3
If the AP is not getting an IP address, this indicates a Layer 2 issue between the DHCP server and the AP.

This is typically cased by a VLAN mismatch.
APs are discovered by DHCP options
Userlevel 3
How is the AP discovering the controller? DHCP options, or something else?
I've just checked, i misread the LLDP the AP it isn't getting an IP address. tried a reboot several time but nothing.
Userlevel 4
Strange, try do disable and enable PoE to reboot the AP. If this don't work, no idea. Maybe you can try to SSH from the Cisco switch direct, if it's firewall or routing problem.
using LLDP I can see a Chassis ID which show an IP address locally obtained by DHCP I can ping the address but i cannot access via SSH.

I have checked adoption history and I cannot see any info for this AP.
Userlevel 4
Does the AP get a IP address and can you connect? Check for information via LLDP or CDP.

Do you check the adoption history? Any connection from this AP?
It doesn't reach the controller, i'm afraid the AP is about 6 hours drive away now so i'm unable to console on to the AP any other suggestions?
Userlevel 2
Can you connect through the console to the APs?

With this we could confirm if it is HW problem
Userlevel 4
Please see POE output below:

Interface Admin Oper Power Device Class Max
(Watts)
--------- ------ ---------- ------- ------------------- ----- ----
Fa2/0/23 auto on 15.4 Ieee PD 4 15.4

Interface AdminPowerMax AdminConsumption
(Watts) (Watts)
---------- --------------- --------------------

Fa2/0/23 15.4 15.4

show adoption history

With this command you can see why it's not adopted (if it can reach the controller).
Please see POE output below:

Interface Admin Oper Power Device Class Max
(Watts)
--------- ------ ---------- ------- ------------------- ----- ----
Fa2/0/23 auto on 15.4 Ieee PD 4 15.4

Interface AdminPowerMax AdminConsumption
(Watts) (Watts)
---------- --------------- --------------------

Fa2/0/23 15.4 15.4

Timo your advice fixed the problem!! (no lldp tlv-select power-management command fixed it ) out of the three APs I have all but one working, so now my issue is the last AP powers up and stays up but it doesn't get adopted i cannot see it getting a DHCP address.

Any advice?
Userlevel 6
Please see POE output below:

Interface Admin Oper Power Device Class Max
(Watts)
--------- ------ ---------- ------- ------------------- ----- ----
Fa2/0/23 auto on 15.4 Ieee PD 4 15.4

Interface AdminPowerMax AdminConsumption
(Watts) (Watts)
---------- --------------- --------------------

Fa2/0/23 15.4 15.4

Timo,

Time-Lenght-Value (TLV) extension carries Power via Media-Dependent-Interface (MDI). See snapshot from my lab



If the power budged is too low AP should not boot correctly as PSE allocated value won't fit classification signature at AP side.

Disabling the LLDP on each side will bring no negotiation, hence full AF power.
Which seems to be provided anyway.

Michale could capture LLDP on switch side to see what is AP asking for, but I am afraid it won't bring much more light. If other switches / ports and cables were tested the fault seems to be at AP side.

Regards,
Ondrej
Userlevel 4
Please see POE output below:

Interface Admin Oper Power Device Class Max
(Watts)
--------- ------ ---------- ------- ------------------- ----- ----
Fa2/0/23 auto on 15.4 Ieee PD 4 15.4

Interface AdminPowerMax AdminConsumption
(Watts) (Watts)
---------- --------------- --------------------

Fa2/0/23 15.4 15.4

Looks like a power handling problem? I think it was LLDP-MED that negotiate the power levels between switch and ap. Try to check this settings and disable it.

Is it possible that the power budget is to low, if the APs get 15.4 Watt?
Userlevel 6
Please see POE output below:

Interface Admin Oper Power Device Class Max
(Watts)
--------- ------ ---------- ------- ------------------- ----- ----
Fa2/0/23 auto on 15.4 Ieee PD 4 15.4

Interface AdminPowerMax AdminConsumption
(Watts) (Watts)
---------- --------------- --------------------

Fa2/0/23 15.4 15.4

3af is more than enough to keep these APs alive.
My AP7522 lives well, serving MCX, with as high as 7.8mW



Regards,
Ondrej
Please see POE output below:

Interface Admin Oper Power Device Class Max
(Watts)
--------- ------ ---------- ------- ------------------- ----- ----
Fa2/0/23 auto on 15.4 Ieee PD 4 15.4

Interface AdminPowerMax AdminConsumption
(Watts) (Watts)
---------- --------------- --------------------

Fa2/0/23 15.4 15.4
Userlevel 4
Can you check, how many watt the APs use and what class the APs report?

Reply