Clearing SEL Fullness via Redfish API

[nsargent]

Hello, I am wondering if anyone can direct me to any additional information on exactly how to process an http request with any specific parameters that will clear SEL Fullness in the Cisco IMC. I have attempted to use documented steps with LogService.ClearLog but it does not appear to process, and the logs do not clear. Per the Developer's page this is what I have attempted to process:

curl -GET https://10.10.10.10/redfish/v1/Chassis/Server1/LogServices/SEL/Actions/LogService.ClearLog -d '{}' -u admin:XXXXXXXXX --insecure -v

Response:

*   Trying 10.10.10.10:443...

* Connected to 10.10.10.10 (10.10.10.10) port 443 (#0)

* ALPN: offers h2,http/1.1

* could not load PEM client certificate from T, OpenSSL error error:80000002:system library::No such file or directory, (no key found, wrong pass phrase, or wrong file format?)

* Closing connection 0

curl: (58) could not load PEM client certificate from T, OpenSSL error error:80000002:system library::No such file or directory, (no key found, wrong pass phrase, or wrong file format?)

Fault Entries:

curl -k -u admin:XXXXXXXXX https://10.10.10.10/redfish/v1/Managers/CIMC/LogServices/Fault/Entries

{

  "@odata.count":"1",

  "Members":[{

      "Name":"Log Entry 1",

      "Id":1,

      "MessageId":"F0462",

      "Message":"SEL_FULLNESS: System Event log is Full: Clear the log",

      "EntryType":"Fault",

      "Description":"Log Entry 1",

      "Created":"2022-10-04T08:30:48",

      "@odata.id":"/redfish/v1/Managers/CIMC/LogServices/Fault/Entries/1",

      "Severity":"1"

    }],

  Members@odata.count:1,

  "@odata.type":"#LogEntry.1.0.0.LogEntryCollection",

  "@odata.id":"/redfish/v1/Managers/CIMC/LogServices/Fault/Entries",

  "@odata.context":"/redfish/v1/$metadata#Managers/Members/$entity/LogServices/Members/$entity/Entries"

[Brian Morrissey]

On a C220 M5 running 4.2(3b) this worked for me:

curl -k -u admin https://cimc_ip/redfish/v1/Managers/CIMC/LogServices/SEL/Actions/LogService.ClearLog -d '{}'

I think what happened with yours is when you use curl with -GET it's combining the flags -G (get) and -E (use client certificate) and T (name of certificate file on your machine)

[nsargent]

Hi Brian,

Thanks for the response. I see the response does go through, and I have experienced this before, but the logs never actually get cleared. Here is the end of the verbose response:

>
< HTTP/1.1 200
< Server: nginx
< Date: Thu, 13 Jul 2023 12:26:07 GMT
< Content-Type: application/json
< Content-Length: 0
< Connection: keep-alive
<
* Excess found: excess = 1 url = /redfish/v1/Managers/CIMC/LogServices/SEL/Actions/LogService.ClearLog (zero-length body)

[Brian Morrissey]

Do you have the server model and firmware version handy?  I've seen some differences between m4 and m5 generations or with platforms like the s3260s

[nsargent]

UCS s3260
I believe it's an m4. And the firmware is 4.1(3d), although I did look via Redfish and it's saying 3.0(4j) and it's saying that on all my servers which can't be right.

[nsargent]

i stand corrected. it is actually 3.0(4j) ... though the m5's are on 4.1(3d) evidently.

[nsargent · Accepted solution]

Okay for whatever reason, the path:

/redfish/v1/Managers/CIMC/LogServices/SEL/Actions/LogService.ClearLog

does not work, but I was able to execute the clear log under:

redfish/v1/Chassis/Server1/LogServices/SEL/Actions/LogService.ClearLog

The response still had a 0 length header with a 1 byte response (the excess noted at the end).

[Brian Morrissey]

Just to confirm the 2nd endpoint worked and cleared the SEL log?  Or it still accepts the request but doesn't actually clear the SEL log?  If it is still not working the other thing you can try is a different s3260 specific endpoint '/redfish/v1/Chassis/CMC/LogServices/SEL/Actions/LogService.ClearLog' to see if that makes any difference. 

If its still not clearing with valid redfish commands it sounds like TAC may need to investigate issues with the firmware

[nsargent]

Yes, the 2nd endpoint ended up working for this M4.