[Starlingx-discuss] Question about alarms in StarlingX

Waines, Greg Greg.Waines at windriver.com
Sun Jun 18 10:47:55 UTC 2023


Ah ok …

Yeah in StarlingX / Kubernetes environment, it is the StarlingX Infrastructure / Flock Services that is responsible for management of the baremetal servers and platform software running on it.

So it would make sense for the software that you are proposing to use StarlingX FM APIs to generate these alarms/logs.  They would then be visible to the end user thru StarlingX Horizon GUI, StarlingX FM CLI/API and optionally SNMP.



Although, you’re probably aware that StarlingX does support running Linux Auditd which also detects and logs security violation events … with configurable auditd.rules.  The logs by default go to  /var/log/audit/audit.log …. i.e. they are not currently forwarded to StarlingX FM.

Does the software that you are thinking of contributing provide different capability than Linux Auditd ?

Greg.


From: Richard Schmitt <richard.schmitt at starlab.io>
Sent: Friday, June 16, 2023 10:11 AM
To: Waines, Greg <Greg.Waines at windriver.com>; starlingx-discuss at lists.starlingx.io
Subject: Re: Question about alarms in StarlingX

Thanks for the detailed answer.

Let me answer the question, “What software specifically do you want to generate alarms?”

We have software that monitors security events associated with the Linux kernel that we would like to expose to the cloud platform.  These security events get logged to journald which in turn is picked up by syslog.  We would make a syslog rule that selects these specific audit events and outputs them to a named pipe.  We’d run a service on each node that would read from this pipe, massage the syslog message, and, our hope, would be to create an alarm within StarlingX that represents that security event.

That’s really the crux of my question.  I’ve read that Kubernetes generally does not handle platform events yet we are exploring a requirement that these security alarms be exposed to the cloud admin through the Horizon web GUI.  I’m trying to understand if this is possible.

Again, thanks for taking the time to read my question and help.

Rich

From: Waines, Greg <Greg.Waines at windriver.com<mailto:Greg.Waines at windriver.com>>
Date: Thursday, June 15, 2023 at 8:19 PM
To: Richard Schmitt <richard.schmitt at starlab.io<mailto:richard.schmitt at starlab.io>>, starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io> <starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>>
Subject: RE: Question about alarms in StarlingX
[EXTERNAL SENDER]
Hey Richard … sorry for the slow reply …
This is the right place to ask.

I am not totally sure what you mean by “We want to create user alarms in OpenStack.”

  *   What software specifically do you want to generate alarms related to what specific fault conditions ?

Some other background on alarms in StarlingX:

  *   StarlingX Platform has a Fault Management system of its own

     *   It is used by StarlingX Platform’s Infrastructure Services to raise alarms and logs related to fault conditions detected by StarlingX Platform’s Infrastructure Services
     *   These alarms / logs are stored in a historical event (alarm or log) db
     *   And an active alarm db is maintained to show alarms that are currently present within StarlingX
     *   The historical event db and the active alarm db can be displayed thru

        *   the starlingx API/CLI ‘fm’, and
        *   the starlingx horizon, on the Fault Management page

     *   There is an internal developer’s API for FM for people coding StarlingX Platform Infrastructure Services code to raise and clear alarms/logs

        *   Likely not documented well … other than in the code (lol)
        *   Let me know if this is what you are looking for

  *   StarlingX OpenStack … i.e. the containerized openstack solution

     *   You’re right that the only pure openstack alarming that I am aware of is via aodh which does alarm thresholding of statistics collected by openstack telemetry / gnoochi … which StarlingX OpenStack doesn’t fully support anymore … we let it go stale and need to fix it
     *   …
     *   However there is a StarlingX OpenStack VIM component that lives in the gray area between the StarlingX Platform on the host and StarlingX OpenStack in containers …

        *   One of the things VIM does, is watch openstack events and raises FM alarms and logs for many openstack events
        *   I believe this is still operational and working today

     *   At one time we were working on cloning FM and having a StarlingX Platform FM and a StarlingX OpenStack FM … but I don’t believe that ever got completed.

Anyways … let me know if this helps,
Greg


From: Richard Schmitt <richard.schmitt at starlab.io<mailto:richard.schmitt at starlab.io>>
Sent: Monday, June 12, 2023 5:59 PM
To: starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: [Starlingx-discuss] Question about alarms in StarlingX

Sorry if this isn’t the right place to ask this question.

We want to create user alarms in OpenStack.  We are not very knowledgeable of OpenStack.  Ultimately, we’d like to raise alarms and have them appear in the Horizon Fault Management page or via the fm CLI.  It does not appear that StarlingX has anything to do with creating or raising alarms.  There are some references in OpenStack documentation on aodh but that service is not deployed in our configuration.

I’m not entirely sure whether OpenStack supports extensible alarms.  Perhaps that’s why I’m finding so little info on it.  Any pointers on where I should look?

Thanks,
Rich


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.starlingx.io/pipermail/starlingx-discuss/attachments/20230618/5baae6ab/attachment-0001.htm>


More information about the Starlingx-discuss mailing list