[phosphor-host-ipmid] Questions about IPMI event message

[phosphor-host-ipmid] Questions about IPMI event message

[Chris Chen (TPI)]

[-- Attachment #1: Type: text/plain, Size: 2683 bytes --]

Hi all,

I sent an IPMI command for testing the event message feature, the command and response as below, looks the event has been sent.

======
$ ipmitool -I lanplus -H 192.168.1.20 -C 17 -U root -P 0penBmc event 1
Sending SAMPLE event: Temperature - Upper Critical - Going High
Get SDR 0004 command failed: Invalid data field in request
Get SDR 0004 command failed: Invalid data field in request
Get SDR 0004 command failed: Invalid data field in request
Get SDR 0004 command failed: Invalid data field in request
Get SDR 0004 command failed: Invalid data field in request
   0 |  Pre-Init  |0000000000| Temperature #0x30 | Upper Critical going high | Asserted
======

Then, I checked the result on the platform, I am able to see the enevt has been added in the "/var/log/redfish" (of course I also can see it on WebUI), however, it doesn't appear in the "var/log/ipmi_sel" or the result of using "ipmitool sel elist".

======
root@hudsonbay-obmc:~# cat /var/log/redfish
1970-01-01T00:00:13.052207+00:00 OpenBMC.0.1.DCPowerOn,
1970-01-01T00:00:15.380591+00:00 OpenBMC.0.1.SELEntryAdded,2C0408666F00FFFF
1970-01-01T00:00:23.481564+00:00 OpenBMC.0.1.InventoryAdded,hudsonbay,Board,sssnnn-12312312
1970-01-01T00:00:32.281467+00:00 OpenBMC.0.1.SELEntryAdded,2C04218008F10C0A
1970-01-01T00:01:47.572079+00:00 OpenBMC.0.1.SELEntryAdded,2C04218008F00C0A
1970-01-01T00:03:42.089885+00:00 OpenBMC.0.1.SELEntryAdded,2C0401300109FFFF
root@hudsonbay-obmc:~#
root@hudsonbay-obmc:~#
root@hudsonbay-obmc:~# ipmitool sel elist
SEL has no entries
root@hudsonbay-obmc:~#
======

So, I have 2 questions for this and hope someone who had experience with this could share ideas with me, I'll appreciate it:

  1.  Why the event doesn't appear in the "var/log/ipmi_sel" or can't be seen by using "ipmitool sel elist"? I expected it should appear.
  2.  What does "Get SDR 0004 command failed: Invalid data field in request" mean?

Things to Note:

  1.  The version of OprnBMC codebase I used is 2.10-rc1.
  2.  The image I built included "intel-ipmi-oem", "phosphor-logging", "phosphor-sel-logger", "rsyslog" and "entity-manager".

Thank you.

Regards,
Chris Chen

Legal Disclaimer :
The information contained in this message may be privileged and confidential. 
It is intended to be read only by the individual or entity to whom it is addressed 
or by their designee. If the reader of this message is not the intended recipient, 
you are on notice that any distribution of this message, in any form, 
is strictly prohibited. If you have received this message in error, 
please immediately notify the sender and delete or destroy any copy of this message!

[-- Attachment #2: Type: text/html, Size: 5839 bytes --]

Re: [phosphor-host-ipmid] Questions about IPMI event message

[Andrei Kartashev]

Hi Chris,

> So, I have 2 questions for this and hope someone who had experience
> with this could share ideas with me, I'll appreciate it:
>    1. Why the event doesn't appear in the "var/log/ipmi_sel" or can't
> be seen by using "ipmitool sel elist"? I expected it should appear.

Disclaimer: here and below I talk only about file-based implementation
for logs.

It is known limitation. First of all, you should aware that there are
two completely different logs: Redfish Log and SEL log. When an
internal sensor triggers threshold event, two different messages are
sent: one for redfish log and another to SEL.

The sel-logger currently designed to only handle threshold events from
internal sensors, thus it requires dbus path to the sensor as an
argument for IpmiSelAdd. 
When you send external IPMI event/SEL add message, ipmi daemon doesn't
know how to handle it. Instead, intel-ipmi-oem, tries to parse the
event according to some rules and if it fails just adds this weird
"OpenBMC.0.1.SELEntryAdded,2C0401300109FFFF" to the redfish log, saying
"we got some ipmi event, but we don't know what is it".

I extended sel-logger with AddExt method, to store any random SEL entry
to /var/log/ipmi_sel, but this patches are for our internal Intel-BMC
fork and not applied for upstream version.
It is on my roadmap to update patches and submit them. If there is
demand for this work, I can try to prioritize it.

>    2. What does "Get SDR 0004 command failed: Invalid data field in
> request" mean?
> 
> Things to Note: 
>    1. The version of OprnBMC codebase I used is 2.10-rc1.
>    2. The image I built included "intel-ipmi-oem", "phosphor-logging",
> "phosphor-sel-logger", "rsyslog" and "entity-manager".
> 
> Thank you.
> 
> Regards,
> Chris Chen 
> Legal Disclaimer :
> The information contained in this message may be privileged and
> confidential. 
> It is intended to be read only by the individual or entity to whom it
> is addressed 
> or by their designee. If the reader of this message is not the intended
> recipient, 
> is strictly prohibited. If you have received this message in error, 
> please immediately notify the sender and delete or destroy any copy of
> this message!

-- 
Best regards,
Andrei Kartashev

回覆: [phosphor-host-ipmid] Questions about IPMI event message

[Chris Chen (TPI)]

[-- Attachment #1: Type: text/plain, Size: 4532 bytes --]

Hi Andrei,

>>Disclaimer: here and below I talk only about file-based implementation
>>for logs.

>>It is known limitation. First of all, you should aware that there are
>>two completely different logs: Redfish Log and SEL log. When an
>>internal sensor triggers threshold event, two different messages are
>>sent: one for redfish log and another to SEL.

>>The sel-logger currently designed to only handle threshold events from
>>internal sensors, thus it requires dbus path to the sensor as an
>>argument for IpmiSelAdd.
>>When you send external IPMI event/SEL add message, ipmi daemon doesn't
>>know how to handle it. Instead, intel-ipmi-oem, tries to parse the
>>event according to some rules and if it fails just adds this weird
>>"OpenBMC.0.1.SELEntryAdded,2C0401300109FFFF" to the redfish log, saying
>>"we got some ipmi event, but we don't know what is it".
Now I get it. I have to say this explanation help me to understand more details. Thanks a lot.
Another, may I consult you other related questions? That is

  *   I had a question about "sel-logger" before, and Lei suggested I to enable "send-to-logger" option, does this have any relevance you mentioned above?
  *   I saw some companies have the xxx-yaml-config folder with "xxx-ipmi-sensors.yaml" etc. and also have the "entity-manager" folder with "xxx.json". But some companies only have "entity-manager" folder. So I would like to understand what differences they are? Should I need them both? I for now use "entity-manager" only.

>>I extended sel-logger with AddExt method, to store any random SEL entry
>>to /var/log/ipmi_sel, but this patches are for our internal Intel-BMC
>>fork and not applied for upstream version.
>>It is on my roadmap to update patches and submit them. If there is
>>demand for this work, I can try to prioritize it.
It definitely will be helpful for me if you are able to prioritize this feature.:)

Regards,
Chris Chen


________________________________
寄件者: Andrei Kartashev <a.kartashev@yadro.com>
寄件日期: 2021年11月3日 下午 10:06
收件者: Chris Chen (TPI) <Chris.Chen3@flex.com>; openbmc@lists.ozlabs.org <openbmc@lists.ozlabs.org>
主旨: Re: [phosphor-host-ipmid] Questions about IPMI event message


Hi Chris,

> So, I have 2 questions for this and hope someone who had experience
> with this could share ideas with me, I'll appreciate it:
>    1. Why the event doesn't appear in the "var/log/ipmi_sel" or can't
> be seen by using "ipmitool sel elist"? I expected it should appear.

Disclaimer: here and below I talk only about file-based implementation
for logs.

It is known limitation. First of all, you should aware that there are
two completely different logs: Redfish Log and SEL log. When an
internal sensor triggers threshold event, two different messages are
sent: one for redfish log and another to SEL.

The sel-logger currently designed to only handle threshold events from
internal sensors, thus it requires dbus path to the sensor as an
argument for IpmiSelAdd.
When you send external IPMI event/SEL add message, ipmi daemon doesn't
know how to handle it. Instead, intel-ipmi-oem, tries to parse the
event according to some rules and if it fails just adds this weird
"OpenBMC.0.1.SELEntryAdded,2C0401300109FFFF" to the redfish log, saying
"we got some ipmi event, but we don't know what is it".

I extended sel-logger with AddExt method, to store any random SEL entry
to /var/log/ipmi_sel, but this patches are for our internal Intel-BMC
fork and not applied for upstream version.
It is on my roadmap to update patches and submit them. If there is
demand for this work, I can try to prioritize it.

>    2. What does "Get SDR 0004 command failed: Invalid data field in
> request" mean?
>
> Things to Note:
>    1. The version of OprnBMC codebase I used is 2.10-rc1.
>    2. The image I built included "intel-ipmi-oem", "phosphor-logging",
> "phosphor-sel-logger", "rsyslog" and "entity-manager".
>
> Thank you.
>
> Regards,
> Chris Chen
> Legal Disclaimer :
> The information contained in this message may be privileged and
> confidential.
> It is intended to be read only by the individual or entity to whom it
> is addressed
> or by their designee. If the reader of this message is not the intended
> recipient,
> is strictly prohibited. If you have received this message in error,
> please immediately notify the sender and delete or destroy any copy of
> this message!

--
Best regards,
Andrei Kartashev



[-- Attachment #2: Type: text/html, Size: 19245 bytes --]