-
Notifications
You must be signed in to change notification settings - Fork 2
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
FVT1060:Invalid power cap value is set via IPMI. #3639
Comments
From Chris Cain: The IPMI request ends up writing the power cap to the dbus. occ-control gets notified after the cap gets written, and sees that it is out of bounds so it corrects it to a legal value. The supported power limits are on dbus: busctl -l introspect xyz.openbmc_project.Settings /xyz/openbmc_project/control/host0/power_capNAME TYPE SIGNATURE RESULT/VALUE FLAGS occ-control could log an error when it has to modify the cap to be within range, but the IPMI command would still be successful (user would NOT know unless they looked for PELs after the set) I do see the following yaml file which should be establishing some limits for power cap: Not sure how that settings would be enforced (obviously it is currently not enforced). |
@lxwinspur , is this something you or someone on the IPS team could investigate? |
@mzipse Our new motherboard (FP5280G3) has just come back, and we are busy debugging the new motherboard during this period, so we have not paid attention to these issues yet. Could IBM please take a look first? |
@yadlapati |
@lxwinspur , testing found this from our 1060 branch. |
@lxwinspur , after further discussions with Chris Cain, our Power Mgmt/OCC Control expert, we are going to live with this problem. Our thinking is that even if an invalid power cap is set, the firmware will still correct it internal to a legal limit. Also, most of our IBM customers will be using Redfish and the GUI to set the power cap which is handled correctly. If you think this needs to be fixed, I'll let you investigate. But if you are ok that there is no error posted when an invalid power cap is set via IPMI, then you can go ahead and close this issue. |
Hi, @mzipse |
@lxwinspur occ-control currently gets the cap limts directly from the OCC and writes them to the dbus: |
Description
======================
Invalid power cap value is set via IPMI.
System details
=================
System : bonn008(service/0penBmc0)
BMC fw : fw1060.00-4.33-1060.2410.20240308a (NL1060_033)
Steps to re- create
======================
Step1 : Login to system
Step2: Executing the IPMI command to set the power cap value 600 and 500 , Min and Max value is 583 and 3982:
Value 600:
bash-4.2$ ipmitool -I lanplus -C 17 -N 3 -p 623 -U ipmi_admin -H bonn008.aus.stglabs.ibm.com dcmi power set_limit limit 600
Password:
Current Limit State: Power Limit Active
Exception actions: Hard Power Off & Log Event to SEL
Power Limit: 600 Watts
Correction time: 0 milliseconds
Sampling period: 0 seconds
Value 600 is set successfully
Step3:
Value 500:
bash-4.2$ ipmitool -I lanplus -C 17 -N 3 -p 623 -U ipmi_admin -H bonn008.aus.stglabs.ibm.com dcmi power set_limit limit 500
Password:
Current Limit State: Power Limit Active
Exception actions: Hard Power Off & Log Event to SEL
Power Limit: 500 Watts
Correction time: 0 milliseconds
Sampling period: 0 seconds
The Value 500 is below the minimum value ,but its taking the value has 556
Actual behaviour: Able to set a invalid value(below the minimum power cap value ) and with no error response for invalid value its able to set a random value.
Expected Behaviour: Need to get a error for setting a below minimum power cap value .
The text was updated successfully, but these errors were encountered: