Report the status message field to Fleet Server checkins

[joshdover]

We should report an agent's reason for being in the status it's in to Fleet Server rather than only relying on Agent logs. This will make it easier to debug why an agent is in a particular status.

Related changes need to be made to Fleet Server to save this state in the .fleet-agents index.

[joshdover]

At minimum we should get these changes into Agent and Fleet to begin persisting this data into ES.

A great addition would be to also expose this in the UI:

• On the agent details page in the table of metadata we have today
• As a tooltip on the agent status badge in the table view

[joshdover]

@jen-huang It'd also be great to have this reported in Fleet's agent status telemetry in the UI. Maybe counts by unique message with some cap.

[juliaElastic]

@joshdover can I use the existing last_checkin_message agent field for this, that was introduced for this feature?

I found that the Agent v2 branches started using the CheckinRequest.Message field and Agent last_checkin_message field, do we expect the v2 reporting to override this logic or should we store in separate fields?

[joshdover]

Makes total sense, yes let's use that since v2 should be shipping in this release as well.

[ghost]

Hi @juliaElastic,

We have created below 04 test cases for this feature.

• C168347
• C168348
• C168349
• C168350

Please let us know if we are missing anything.
Thanks

[ghost]

Hi Team,

We have executed 04 testcases for this feature under our Fleet Test run at Fleet Team 8.6 New features test plan and found that it's working fine.

Build Details:

Version: 8.6.0 BC5
BUILD: 58693
COMMIT: ed40c16ce9999cc47ad55c11bb097d2e443b31a6
Artifact link: https://staging.elastic.co/8.6.0-8cf9e954/summary-8.6.0.html

Below are the test cases.

• T1872624
• T1872625
• T1872626
• T1872627

Please let us know if we are missing anything.
Thanks