Chassis power error and time

ADAman · ‎08-08-2012

major,F0408,117295,sys/chassis-1/fault-F0408,power-problem,2012-08-08T13:45:12,Power state on chassis 1 is redundancy-failed

what does this mean? Both the power cables are connected, and they have status "Operable" in Chassis Power GUI.

Other than that, no faults.

Is it possible to set the time without using NTP? Do I have to set time from NX-OS command-line?

rmueller@willkie.com · ‎08-08-2012

We have the same issue.

Cisco support found out that it is an "i2c bus issue" between IOMs, PSUs and fans. The fix is to reseat all PSUs (keep them out for a minute, one after each other with enough time to come up again) and also to reseat the IOMs. The time here is seven minutes until the IOMs are back in operation. The entire process should be non-disruptive but a maintenance window is recommened.

We have not applied the fix yet but I will keep you posted.

padramas · ‎08-08-2012

Hello Atle,

As Robert suggested, fault looks like cosmetic one due to I2C issues. Please open a TAC service request to further investigate it.

Regarding time configuration, you can manually configure it without the need of NTP server.

http://www.cisco.com/en/US/docs/unified_computing/ucs/sw/cli/config/guide/2.0/b_UCSM_CLI_Configuration_Guide_2_0_chapter_0100000.html#task_7D02AB9BD9F441568C782408B87F3FAA

Padma

David Alpizar · ‎08-08-2012

Hi Atle,

I have seen that kind of issues here and there is something you can try in order to correct this problem:

A workaround to clear the I2C faults is to reseat fan trays, PSUs, and IOMs. Here is how to do it:

1. Reseat the fans from fan 1 to fan 8.

2. Reseat the PSUs from PSU 1 to PSU 4.

Pull the components out one at a time from 1 to the last, leave it out for at least 3 minutes before reinsert it back, observe the light to make sure all return to green before pulling the next one.

3. If the above does not work, pull out the IOM that is

flashing amber. Leave it out for 5 minutes and reinsert. Make sure

everything comes up and the lights turn green.

4. Monitor the system while reseating components by issuing

# show tech-support chassis 1 iom 1 brief

5. May have to reseat the second IOM if amber lights shifted.

6. If reseating the IOM does not solve the problem, we may

need to reacknowledge the chassis. This is disruptive to network

connections of the blades. You need a set a downtime to do it.

7. Repeat these steps on all chassis that are having amber lights.

I strongly recommend you upgrade the firmware to 1.4(3q) and above when

you are able to set a downtime. The errors will need to be clear before

the upgrade. Otherwise the message may remain until you reseat the

components.

This is just a cosmetic issue on the UCS environment.

Hope this helps!

David

ADAman · ‎08-09-2012

I no got similar errors. I have 2 serverports configured and the chassis discovery policy is set at 2.

ack="no"

cause="unexpected-number-of-links"

changeSet=""

code="F0440"

created="2012-08-08T15:33:08"

descr="Chassis discovery policy conflict: Link IOM 1/1/1 to fabric interconnect A:1/19 not configured"

dn="sys/chassis-1/slot-1/fabric/port-1/fault-F0440"

highestSeverity="info"

id="117166"

lastTransition="2012-08-09T13:19:56"

lc=""

occur="6"

origSeverity="info"

prevSeverity="cleared"

rule="ether-switch-intfio-satellite-wiring-numbers-unexpected"

severity="info"

status="created"

tags="network"

type="connectivity">

David Alpizar · ‎08-09-2012

Hi Atle ,

I did some research about this error and you can try the following:

- Make sure that number of present IOM uplinks are at least same as that defined in chassis discovery policy link-action. Configure corresponding sever ports.

- Re-acknowledge the chassis to acknowledge the mismatched number of present links.

You can SSH the server and check the alerts using the following commands:

UCS-A# scope chassis X
UCS-A /chassis # show fault

Hope this help.

David

ADAman · ‎08-09-2012

the numbe rof links error is gone. But the original error is still there. I just reseated the two poer supplies. I guess I have to reseat IO momdules now.

ADAman · ‎08-09-2012

this is my status concerning thir error:

FI6120XP-A /chassis # show fault

Severity Code Last Transition Time ID Description

--------- -------- ------------------------ -------- -----------

Cleared F0528 2012-08-09T16:39:57.618 150705 Power supply 2 in chassis 1 power: off

Cleared F0378 2012-08-09T16:39:43.064 150704 Power supply 2 in chassis 1 presence: missing

Cleared F0378 2012-08-09T16:36:24.057 150702 Power supply 1 in chassis 1 presence: missing

Warning F0378 2012-08-08T15:45:12.955 117296 Power supply 3 in chassis 1 presence: missing

Warning F0378 2012-08-08T15:45:12.955 117297 Power supply 4 in chassis 1 presence: missing

Major F0408 2012-08-08T15:45:12.954 117295 Power state on chassis 1 is redundancy-failed

So it is still there. I will now reseat the IOMs.

David Alpizar · ‎08-09-2012

Atle ,

I can see that the link error is cleared now.

Now, talking about the power supply issue, try to reseat the IOMs. The alerts should be gone, but if they are still there you can open a TAC case.

You can have a faulty power supply on the chassis.

David

ADAman · ‎08-09-2012

yes, reseating IOMS did not help