core: Fix standalone orchestrator not crashing under UnrecoverableError

[leszko]

What does this pull request do? Explain your changes. (required)

Fix the issue that O does not fail after the unrecoverable error from lpms (ffmpeg).

The issue is described in this Discord thread. In short, when GPU crashes, NVIDIA recommends restarting the whole process, since it's not possible to recover. It worked that way until the Gracefully notify orchestrator in case of a panic in transcoder PR. In the mentioned PR, the following logic was implemented:

1. Recover from LPMS (ffmpeg) panic
2. Send a message T->O
3. Panic

The logic is fine for the split O/T topology, however for the combined OT, there is a separate code which also needs to implement Point 3. Panic.

Specific updates (required)

How did you test each of these updates (required)

Artificially introduce panic() and check that a standalone O crashes.

Does this pull request close any open issues?

Checklist:

• Read the contribution guide
• make runs successfully
• All tests in ./test.sh pass
• README and other documentation updated
• Pending changelog updated

[victorges]

Nice! Just a few questions/discussions on the crashing logic

[victorges]

LGTM