[python] test_events: ignore Unavailable status events - #6799

test_events asserts an exact status-event sequence, so it fails whenever the runner
happens to record Unavailable. That status means the runner momentarily could not reach
the pipeline, which reflects the cluster rather than the lifecycle under test.

Describe Manual Test Plan

Seen in CI a compiler-server outage crash-looped the pipeline container, the pod left service, its DNS
record was withdrawn, and the runner recorded one Unavailable event between Initializing
and Running:

At index 5 diff: ('Provisioned', 'Provisioned', 'Unavailable', 'Unavailable', False, 'Success', 'InUse')
             != ('Provisioned', 'Provisioned', 'Initializing', 'Running', False, 'Success', 'InUse')

Replayed the 47 events that run recorded (from its support bundle) through the test's
mapping code:

The last row matters: the filter drops only Unavailable, and the test still catches a
lifecycle transition that goes missing.

Checklist

• Unit tests added/updated
• Integration tests added/updated
• Documentation updated
• Changelog updated

Breaking Changes?

None.