iLO Amplifier Pack -- Server Refresh Job failed

[sergiopt]

I am trying to automate a server refresh job after a firmware update using the iLO Amplifier Pack REST API. The process works if I start the server refresh through the website, but when doing it through the REST API I get the following error message on most servers: 

"Server Refresh job failed because the Server inventory details not found in the iLO Amplifier. "

I am using the endpoint https://{iLOAmpServer}/redfish/v1/JobService/Actions/Oem/Hpe/HpeWfmJobServiceExt.RefreshJob/ with the following JSON body:

{

            "SelectedSystemsManagerAddress": [

                        "{Manager1_IP_Address}",

                        "{Manager2_IP_Address}"

            ]

}

The issue doesn't seem to be related to missing permissions, since I can execute the server refresh job through the website using the same user credentials. If I do this and then execute the REST API call again, sometimes the server refresh job will work afterwards.

[support_s]

System recommended content:

1. HPE iLO 5 2.98 User Guide | iLO Amplifier Pack

2. HPE iLO 4 2.82 User Guide | iLO Amplifier Pack

Please click on "Thumbs Up/Kudo" icon to give a "Kudo".

Thank you for being a HPE valuable community member.

[Kashyap02]

Hi, 

I tried in the our lab, it works. I run the below POST method. 

POST https://<Amplifier IP>/redfish/v1/JobService/Actions/Oem/Hpe/HpeWfmJobServiceExt.RefreshJob

With below details in Body: 

{
"SelectedSystemsManagerAddress": 
[
"<Server 1 ILO IP>",
"<Server 2 ILO IP>"
]
}

Output was like below. 

{
  "@odata.etag": "W/\"<ID>\"",
  "@odata.type": "#ExtendedInfo.1.0.0.ExtendedInfo",
  "error": {
    "@Message.ExtendedInfo": [{
      "@odata.type": "#Message.1.0.0.Message",
      "MessageArgs": ["/redfish/v1/JobService/Jobs/10011"],
      "MessageId": "HpeCommon.2.0.JobCreated"
    }],
    "code": "Wolfram.1.0.ExtendedInfo",
    "message": "See @Message.ExtendedInfo for more information."
  }
}

Even it showed the error, the refresh job was executed successfully. 

The job was started immidiately and completed. 



I noticed one thing where the refresh job was run on the server which was in unknown state in Amplifier.. Resulted in Exception. The message was "Server Refresh job failed because credentials used for server are invalid."

The method and the steps you are trying is correct. 
Check the Amplifier version and ILO firmware as well.

[sergiopt]

Hi, 

thanks for the reply.

The problem occurs only when starting the server refresh job through the REST API. The job is successfully created but it fails after 35% showing the following error message:

"Server Refresh job failed because the Server inventory details not found in the iLO Amplifier. "

If I manually start a refresh server job on the web interface seconds after, it completes without showing any error message:

The automatic server refresh which is scheduled to run every hour also completes without issues. The problem seem to be related to the way the REST API call is handled on the server.

The iLO Amplifier Pack is up to date (Version 2.22) and I have tried the refresh job on an iLO 5 with both firmware 2.91 and 2.97. The error appears on both firmware versions.

[Kashyap02]

Hi, 

You can try clearning the RESTAPI state in ILO. There is a chance of the stale data which might be causing this. 

Login into ILO CLI using Administrator credentials. 
Issue the command oemhpe_clearRESTAPIstate
Run the refresh Job from using REST API.. 

Note: The ILO will reset after execuring the above command. This will not have any impact on the production.

Try this on one server. 

[sergiopt]

Unfortunately clearing the RESTAPI state in ILO hasn't solved the issue either. The server refresh fails showing the same error message when triggered through the REST API call but works without issues when I trigger it through the iLO Amplifier Pack web interface.

[Kashyap02 · Accepted solution]

Hi, 

Is this issue noticed on all the servers or one server. If one server refresh is failing, you can remove the server from Amplifier and re-add it. If this is happening across all the servers, recommend you to open a formal ticket with HPE. 

[sergiopt]

Hello,

thanks for the suggestion. The server refresh is now working after deleting the server from the iLO Amplifier Pack and adding it again.

This was happening on several servers until a few days ago. I've been trying to reproduce it, but it seems to be working now.

I will open a ticket with HPE if I encounter this issue again.

Thank you very much.

[Sunitha_Mod]

Hello @sergiopt,

Perfect! 

We are extremely glad to know the problem has been resolved and we appreciate you for keeping us posted.