8.16.0-SNAPSHOT: standalone to managed migration is broken

[marclop]

Description

The migration of apm-server from standalone to managed mode is broken.

Error: error loading config file: stat apm-server.yml: no such file or directory

Reproduce

Manual

1. Create APM Server 7.17.25 standalone
2. Upgrade to 8.16.0-SNAPSHOT
3. Migrate from standalone to managed mode.
4. APM Server will break

Smoke test

$ make smoketest/run-version TEST_DIR=./testing/smoke/legacy-standalone-major-managed SMOKETEST_VERSION=8.16.0-SNAPSHOT
-> Running ./testing/smoke/legacy-standalone-major-managed smoke tests for version latest...
-> Running 7.17.25-SNAPSHOT standalone to 8.16.0-SNAPSHOT to 8.16.0-SNAPSHOT managed
-> Adding tfvar stack_version="7.17.25-SNAPSHOT"
-> Adding tfvar integrations_server="false"
-> Applying terraform configuration...
-> APM Server ready!
-> Sending events to APM Server...
-> Asserting legacy ingest pipelines...
-> Asserting legacy index templates...
-> Asserting legacy ILM policies...
-> Asserting apm-7.17.25-error-* contains expected documents...
-> Asserted 1 error.id=9876543210abcdeffedcba0123456789 exists
-> Asserting apm-7.17.25-span-* contains expected documents...
-> Asserted 1 span.id=1234567890aaaade exists
-> Asserting apm-7.17.25-transaction-* contains expected documents...
-> Asserted 1 transaction.id=4340a8e0df1906ecbfa9 exists
-> Asserting apm-7.17.25-metric-* contains expected documents...
-> Asserted 1 transaction.type=request exists
-> Adding tfvar stack_version="8.16.0-SNAPSHOT"
-> Adding tfvar integrations_server="false"
-> Applying terraform configuration...
-> APM Server ready!
-> Sending events to APM Server...
-> Asserting ingest pipelines...
-> Asserting logs-apm.error-* contains expected documents...
-> Asserted 1 error.id=9876543210abcdeffedcba0123456789 exists
-> Asserting traces-apm-* contains expected documents...
-> Asserted 1 span.id=1234567890aaaade exists
-> Asserting traces-apm-* contains expected documents...
-> Asserted 1 transaction.id=4340a8e0df1906ecbfa9 exists
-> Asserting metrics-apm.internal-* contains expected documents...
-> Asserted 1 transaction.type=request exists
-> Upgrading APM Server 8.16.0-SNAPSHOT to managed mode...
-> Waiting for 90 seconds for the APM Server to become available...
{"ok":false,"message":"The requested resource is currently unavailable."}
-> APM Server isn't ready to receive events, retrying (1/10)...
{"ok":false,"message":"The requested resource is currently unavailable."}
-> APM Server isn't ready to receive events, retrying (2/10)...
{"ok":false,"message":"The requested resource is currently unavailable."}
-> APM Server isn't ready to receive events, retrying (3/10)...
{"ok":false,"message":"The requested resource is currently unavailable."}
-> APM Server isn't ready to receive events, retrying (4/10)...
{"ok":false,"message":"The requested resource is currently unavailable."}
-> APM Server isn't ready to receive events, retrying (5/10)...
{"ok":false,"message":"The requested resource is currently unavailable."}
-> APM Server isn't ready to receive events, retrying (6/10)...
{"ok":false,"message":"The requested resource is currently unavailable."}
-> APM Server isn't ready to receive events, retrying (7/10)...
{"ok":false,"message":"The requested resource is currently unavailable."}
-> APM Server isn't ready to receive events, retrying (8/10)...
{"ok":false,"message":"The requested resource is currently unavailable."}
-> APM Server isn't ready to receive events, retrying (9/10)...
{"ok":false,"message":"The requested resource is currently unavailable."}
-> APM Server isn't ready to receive events, maximum retries exceeded
make: *** [smoketest/run-version] Error 1

[marclop]

Discovered when working on fixing the smoke tests: #14176

[marclop]

This was already reported in elastic/elastic-agent#5615, and it seems like the issue has been identified. I will leave open until it's fixed.

[marclop]

This has now been resolved.