05-02-2012 11:31 PM - edited 03-07-2019 06:28 AM
Hi,
We have a strange problem with a Catalyst 4503 switch.
The switch has a Supervisor II+TS linecard, two PoE users linecards (WS-X4548-GB-RJ45V), and two power supplies in redundancy mode (PWR-C45-1300ACV).
We have about 30 IP phones in the switch (most Cisco IP Phone 6921, and some 8961).
Everything works fine, but if we disconnect one phone and connect it back, then we have the following log:
.....
.....
Apr 30 16:54:06.873 ES: %LINEPROTO-5-UPDOWN: Line protocol on Interface GigabitEthernet2/18, changed state to down
Apr 30 16:54:07.873 ES: %LINK-3-UPDOWN: Interface GigabitEthernet2/18, changed state to down
Apr 30 16:54:07.873 ES: %SWITCH_QOS_TB-5-TRUST_DEVICE_LOST: cisco-phone no longer detected on port Gi2/18, operational port trust state is now untrusted.
Apr 30 16:54:29.921 ES: %ILPOWER-5-ILPOWER_POWER_DENY: Interface Gi2/18: inline power denied
.....
.....
The same thing happens if we connect a new phone, intead of disconnect a working one (we get the "inline power denied" message log)
If we do a "shutdown" and "no shutdown" in all the ports of the linecard (or if we do a reload of the switch), evething works fine again.
The same situation happens, not matter what user PoE linecard we test with.
We think it is not a problem of lack of power becuase we have quite power left when evering works fine (i.e. after a reload of the switch):
XXXXXXXXXXXX#sho power
Power Fan Inline
Supply Model No Type Status Sensor Status
------ ---------------- --------- ----------- ------- -------
PS1 PWR-C45-1300ACV AC 1300W good good good
PS2 PWR-C45-1300ACV AC 1300W good good good
Power supplies needed by system : 1
Power supplies currently available : 2
Power Summary Maximum
(in Watts) Used Available
---------------------- ---- ---------
System Power (12V) 480 939
Inline Power (-50V) 361 800
Backplane Power (3.3V) 0 0
---------------------- ---- ---------
Total 841 (not to exceed Total Maximum Available = 1300)
Module Inline Power Summary (Watts)
(12V -> -48V on board conversion)
---------------------------------
Maximum
Mod Used Available
--- ---- ---------
1 0 158
--- ---- ---------
Thanks very much.
Solved! Go to Solution.
05-03-2012 12:33 AM
Lot of bugs listed.. this might be relevant... (from the release notes). I would probably try upgrading to see if that helps....
This section lists the resolved caveat in Cisco IOS Release 12.2(25)EWA14:
•On a Catalyst 4500 series switch with either dual 4200W AC power supplies with one or both 220V connections, or a single 4200W AC power supply with two 220V connections, a power supply firmware glitch temporarily resets the status bits. It causes Catalyst 4500 IOS software to receive the false input voltage values (from 220V to 110V) and trigger the normal recovery routine.
You will see one or both of the following error messages:
*Mar 5 11:16:33.663 UTC: %C4K_CHASSIS-3-MIXINVOLTAGEDETECTED: Power supplies in the chassis are receiving different voltage inputs
*Mar 5 11:16:33.663 UTC: %C4K_CHASSIS-3-MIXINPOWERDETECTED: Power supplies in the chassis are of different types (AC/DC) or wattage
You may also see the following message:
%ILPOWER-5-ILPOWER_POWER_DENY: Interface Gi2/33: inline power denied
On ports with POE devices connected.
This glitch temporarily shuts down the affected power supply and causes a loss of power supply redundancy. Power for the data and chassis is decremented and might cause the linecard(s) to shut down. Power for PoE also will decrement, and it could cause PDs to shut down and reset.
Note the following:
–Only units with serial numbers starting from with AZS12200001 and higher do not experience this defect.
–If both power supplies have 110V inputs, they are not affected.
The output current will be lower with both 110V input connections, see
Power Supply Calculator on cisco.com at the URL http://tools.cisco.com/cpc/launch.jsp.
Workaround: None. (CSCso67729)
•A group of 4 ports on a PoE linecard might not recognize the IEEE phone device and might not provide inline power. The same device(s) may work in other ports on the same module; it may work with other ports of similar linecards in other slots on the same chassis.
The group(s) of ports on which the PoE devices do not power up may differ in every instance.
The problem has been observed with the following hardware combinations:
–4500 chassis, WS-X4013+=, WS-X4548-GB-RJ45V= (hw rev 4.0 & 4.1), 12.2(40)SG
–4500-E chassis WS-C45-Sup6-E, WS-X4548-GB-RJ45V= (hw rev 4.0 & 4.1), 12.2(40)SG
Note Cisco IP Phone is not affected as it can be detected via CDP.
Workaround: None. (CSCso29149)
•A Catalyst 4500 switch may experience a reload if you perform two or more
"write memory's via CLI" on the switch.
This occurs with Cisco IOS Releases 12.2(25)EWA13 and earlier releases, 12.2(31)SGA6 and earlier releases, and 12.2(4x)SG releases.
Workaround: Restrict vty access to one session and upgrade to Cisco IOS Release 12.2(25)EWA14 or 12.2(31)SGA7. (CSCso86459)
•On a Catalyst 4500 Supervisor Engine running Cisco IOS Releases 12.2(25)EWA13, or 12.2(31)SGA4 through 12.2(31)SGA6, you might receive the message
"NVRAM Verification Failed" and the running config might not be saved to the NVRAM.
Workaround: Upgrade to Cisco IOS Release 12.2(25)EWA14 or 12.2(31)SGA7. (CSCsq27434)
Here's the whole mess...
http://www.cisco.com/en/US/customer/docs/switches/lan/catalyst4500/release/note/OL_24466.html
05-02-2012 11:54 PM
Can you recreate the condition and do a show interface status | inc
Is the port going errdisable? If so, shutting and no shutting the port will clear the condition.
What version of IOS is your switch running also? It could be a bug...
Can debug further with more info
Good luck!
Andy
05-03-2012 12:07 AM
Hi,
.- No, the port is not going errdisable.
.- The IOS's version is:
...
Cisco IOS Software, Catalyst 4000 L3 Switch Software (cat4000-I9S-M), Version 12.2(25)EWA10, RELEASE SOFTWARE (fc1)
Technical Support: http://www.cisco.com/techsupport
Copyright (c) 1986-2007 by Cisco Systems, Inc.
Compiled Wed 11-Jul-07 00:08 by dchih
Image text-base: 0x10000000, data-base: 0x114BC038
ROM: 12.2(20r)EW1
Dagobah Revision 226, Swamp Revision 4
...
.- I will have to wait to do a new debug (when users stop working).
Thanks very much.
05-03-2012 12:30 AM
Hi,
As from the logs , it can be seen that the interface is now untrusted as the interface goes down. and power gets denied. which indicated the switch is not able to recognise the port as trusted even when you connect back . it can be a bug could you share the sh run int
Regards,
Palani
05-03-2012 01:11 AM
Hi,
This is the output for the show run interface command:
.....
.....
interface GigabitEthernet2/18
switchport access vlan 103
switchport mode access
switchport voice vlan 304
service-policy output autoqos-voip-policy
logging event link-status
logging event trunk-status
qos trust device cisco-phone
qos trust cos
auto qos voip cisco-phone
tx-queue 3
priority high
shape percent 33
spanning-tree portfast
end
.....
.....
The problem happens in whatever interface of the two PoE linecards (all of them are cofigured the same as 2/18)
Thank you.
05-03-2012 12:33 AM
Lot of bugs listed.. this might be relevant... (from the release notes). I would probably try upgrading to see if that helps....
This section lists the resolved caveat in Cisco IOS Release 12.2(25)EWA14:
•On a Catalyst 4500 series switch with either dual 4200W AC power supplies with one or both 220V connections, or a single 4200W AC power supply with two 220V connections, a power supply firmware glitch temporarily resets the status bits. It causes Catalyst 4500 IOS software to receive the false input voltage values (from 220V to 110V) and trigger the normal recovery routine.
You will see one or both of the following error messages:
*Mar 5 11:16:33.663 UTC: %C4K_CHASSIS-3-MIXINVOLTAGEDETECTED: Power supplies in the chassis are receiving different voltage inputs
*Mar 5 11:16:33.663 UTC: %C4K_CHASSIS-3-MIXINPOWERDETECTED: Power supplies in the chassis are of different types (AC/DC) or wattage
You may also see the following message:
%ILPOWER-5-ILPOWER_POWER_DENY: Interface Gi2/33: inline power denied
On ports with POE devices connected.
This glitch temporarily shuts down the affected power supply and causes a loss of power supply redundancy. Power for the data and chassis is decremented and might cause the linecard(s) to shut down. Power for PoE also will decrement, and it could cause PDs to shut down and reset.
Note the following:
–Only units with serial numbers starting from with AZS12200001 and higher do not experience this defect.
–If both power supplies have 110V inputs, they are not affected.
The output current will be lower with both 110V input connections, see
Power Supply Calculator on cisco.com at the URL http://tools.cisco.com/cpc/launch.jsp.
Workaround: None. (CSCso67729)
•A group of 4 ports on a PoE linecard might not recognize the IEEE phone device and might not provide inline power. The same device(s) may work in other ports on the same module; it may work with other ports of similar linecards in other slots on the same chassis.
The group(s) of ports on which the PoE devices do not power up may differ in every instance.
The problem has been observed with the following hardware combinations:
–4500 chassis, WS-X4013+=, WS-X4548-GB-RJ45V= (hw rev 4.0 & 4.1), 12.2(40)SG
–4500-E chassis WS-C45-Sup6-E, WS-X4548-GB-RJ45V= (hw rev 4.0 & 4.1), 12.2(40)SG
Note Cisco IP Phone is not affected as it can be detected via CDP.
Workaround: None. (CSCso29149)
•A Catalyst 4500 switch may experience a reload if you perform two or more
"write memory's via CLI" on the switch.
This occurs with Cisco IOS Releases 12.2(25)EWA13 and earlier releases, 12.2(31)SGA6 and earlier releases, and 12.2(4x)SG releases.
Workaround: Restrict vty access to one session and upgrade to Cisco IOS Release 12.2(25)EWA14 or 12.2(31)SGA7. (CSCso86459)
•On a Catalyst 4500 Supervisor Engine running Cisco IOS Releases 12.2(25)EWA13, or 12.2(31)SGA4 through 12.2(31)SGA6, you might receive the message
"NVRAM Verification Failed" and the running config might not be saved to the NVRAM.
Workaround: Upgrade to Cisco IOS Release 12.2(25)EWA14 or 12.2(31)SGA7. (CSCsq27434)
Here's the whole mess...
http://www.cisco.com/en/US/customer/docs/switches/lan/catalyst4500/release/note/OL_24466.html
05-03-2012 01:12 AM
Thank you.
I will probably upgrade the IOS if I can't find another possible solution.
Kind Regards.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide