several H2224XXLR2 systems are throttling, due to one of the PDU's being reported as down, even though there is power on them.

[JanKuipers]

H2224XXLR2 , with S2600TPR board, with most current firmware package R01.01.0028 

The behaviour is identical to the one described in TA-1131 :

"Intel has received numerous reports from customers of unexpected and severe system CPU throttling on the identified products. The event is also coupled with a Power Supply Amber LED warning (1 Hz blink pattern) status generated by the event. A check of the System Event Log (SEL) shows a PSU Predictive Failure, an Over Temperature condition, or a SmART-CLST event."

Is there a more current PSU firmware version? Is this known behaviour? (as it apparantly was before) ?

-Jan

[SergioS_Intel]

Hello JanKuipers,

Please help us proving some additional information about your system configuration:

What processor model are you using?

What's the wattage of your power supply?

What operating system are you running?

Can you please provide us a copy of the System Event Log (SEL) Viewer Utility and the Sysinfo.

https://www.intel.com/content/www/us/en/support/articles/000023940/server-products/server-boards.html

We will be looking forward to your reply.

Best regards,

Sergio S.

Intel Customer Support Technician

[JanKuipers]

dual Intel(R) Xeon(R) CPU E5-2630 v4 @ 2.20GHz

2130W 

XEN/Centos

Attached is the SELLOG, sysinfo is not immediatly available.

-Jan

[SergioS_Intel]

Hello JanKuipers,

When available, please provide us a copy of the sysinfo log, so we can continue assisting you further 

https://www.intel.com/content/www/us/en/support/articles/000023940/server-products/server-boards.html

Best regards,

Sergio S.

Intel Customer Support Technician

[JanKuipers]

posted a file.

[JanKuipers]

posted a file.

[JanKuipers]

posted a file.

[SergioS_Intel]

Hello JanKuipers,

We appreciate the additional information, we are going to check the logs information and will get back to you.

Best regards,

Sergio S.

Intel Customer Support Technician

[Emeth_O_Intel]

Hello JanKuipers,

Thank you so much for been waiting for our response, I am contacting you on behalf of Sergio.

I would like to ask you the following details in order to clarify some details:

1. What is the status of the PSU LED for each PSU? (amber means there is something wrong)
2. Have you performed the Firmware Update using the existing PSU? This is an important procedure as the system update package includes a firmware update for the PSU.

As additional information, based on the sysinfo log it appears to be there is only one PSU in the system. You could consider adding a second one for additional power load.

Also, I will attach the TA-1131.pdf. Please let me know the outcome in order to verify the information.

Emeth O.

Intel® Server Specialist.

[Emeth_O_Intel]

posted a file.

[JanKuipers]

Hi Emeth,

I am aware of TA-1131, and the system has been updated

Despite what the syslog may say, the system does have 2 PSUs connected.

The firmware package update R01.01.0028 (Latest) was applied on 01/09/2020 , which should contain said fixes in TA-1131

However, i see in the bmc->configuration->sdr that the PSU's are _detected_ but the model can't be determined or the FRU data is invalid.

Would that have stopped said fix from TA-1131 to be applied ? Repeated reseating of the PSU's shows no change, nor does total poweroff.

-Jan

[Emeth_O_Intel]

Hi,

Thank you so much for your answer.

I would like to confirm the questions above about your system:

" 

1. What is the status of the PSU LED for each PSU? (amber means there is something wrong)
2. Have you performed the Firmware Update using the existing PSU? This is an important procedure as the system update package includes a firmware update for the PSU.

"

Please let me know the information and I will be more than happy to proceed with the next step.

Best regards,

Emeth O.

Intel® Server Specialist.

[JanKuipers]

1) the amber light is lit.

2) the firmware updates were applied using the existing PSU

-Jan

[SergioS_Intel]

Hello JanKuipers,

We appreciate the additional information. We will get back to you as soon as possible.

Best regards,

Sergio S.

Intel Customer Support Technician

[SergioS_Intel]

Hello JanKuipers,

Based on your response there is something wrong with that PSU.

An amber LED indicates a PSU fault.

The normal status is a green LED: Solid for active PSU, flashing for standby PSU in a redundant configuration.

In this case, since you are using dual PSU configuration, what does the other PSU LED look like?

Can you try swapping the PSU with a different one or even swapping the existing PSUs from bays in the chassis?

The idea is to identify is this is simply a bad PSU or a bad PDB in the system.

Best regards,

Sergio S.

Intel Customer Support Technician

[JanKuipers]

Let me clarify :

the status light of the PSU _when the issue occured_ was amber.

the status light of the PSU is _currently_ green, and green flashing

[SergioS_Intel]

Hello JanKuipers,

We appreciate the clarification, just for testing purposes have you tried to swap the power supplies from bays to chassis?

Regards

[JanKuipers]

Yes we have previously.

Let me stress the fact that we have multiple systems, all with the same behaviour.

- Jan

[SergioS_Intel]

Hello JanKuipers,

We appreciate the additional information, we are going to check the logs information and will get back to you.

Best regards,

Sergio S.

Intel Customer Support Technician

[SergioS_Intel]

Hello JanKuipers,

We just need that you please clarify some information, on an earlier post you mentioned that the amber light is lit and initially when opening the thread:

"The event is also coupled with a Power Supply Amber LED warning (1 Hz blink pattern)."

Now, you mentioned that the status light of the PSU is currently green and green flashing.

We just want to clarify that if something changed or perhaps there was a mistake with the information provided.

An amber LED denotes a hardware issue, if power supply units have been replaced, there still could be something wrong with the PDB (power distribution board), but on the other hand, if all is green with the power supplies, then there doesn't appear to be a hardware problem.

Best regards,

Sergio S.

Intel Customer Support Technician

[JanKuipers]

Hi Sergio,

I though it would be obvious, but applying the AC workaround from TA-1131 "a full power cycle for the units experiencing the issue" returns the led to a green status, and the total system into a non-throttled state.

I would concur that there is no hardware fault.

From march 5th :

The firmware package update R01.01.0028 (Latest) was applied on 01/09/2020 , which should contain said fixes in TA-1131 

 However, i see in the bmc->configuration->sdr that the PSU's are _detected_ but the model can't be determined or the FRU data is invalid.

 Would that have stopped said fix from TA-1131 to be applied ? Repeated reseating of the PSU's shows no change, nor does total poweroff.

--Jan.