feat(k8s): Add pure migration option for api component

[BorisPolonsky]

Important

1. Make sure you have read our contribution guidelines
2. Ensure there is an associated issue and you have been assigned to it
3. Use the correct syntax to link this PR: Fixes #<issue number>.

Summary

This PR ease maintenance in k8s scenario by adding pure migration mode that execute data migration logic and exit normally without launching the web server. This is useful when upgrading dify version, as it enable the maintainer to launch a separate container that execute the flask upgrade-db logic via entrypoint.sh instead of tampering with the actual api or worker

Without this mode, this logic will be executed in both api and worker component instead of a standalone, single replica logic. This will case dead lock in case there are multiple replica of api or worker, or if the worker starts before api and tries to execute the migration logic that end up with issues like #19910. With this migration only mode, it's possible to disable auto migration for api and worker and bring up a standalone container that does the migration beforehand (via a Job defined with pre-installation hook annotation that recognized by Helm).

Screenshots

Before	After
Server will launch after flask upgrade-db	Run flask upgrade-db only without launching server when MODE=migration

Checklist

• This change requires a documentation update, included: Dify Document
• I understand that this PR may be closed in case there was no previous discussion or issues. (This doesn't apply to typos!)
• I've added a test for each change that was introduced, and I tried as much as possible to make a single atomic change.
• I've updated the documentation accordingly.
• I ran dev/reformat(backend) and cd web && npx lint-staged(frontend) to appease the lint gods

[BorisPolonsky]

As alternative solution, we may also change the else statement (at line 29) below and enforce explicit MODE=api condition (i.e. elif [[ "${MODE}" == "api" ]]; then) for the same purpose.

dify/api/docker/entrypoint.sh

Lines 29 to 41 in e9c9c5d

	else
	if [[ "${DEBUG}" == "true" ]]; then
	exec flask run --host=${DIFY_BIND_ADDRESS:-0.0.0.0} --port=${DIFY_PORT:-5001} --debug
	else
	exec gunicorn \
	--bind "${DIFY_BIND_ADDRESS:-0.0.0.0}:${DIFY_PORT:-5001}" \
	--workers ${SERVER_WORKER_AMOUNT:-1} \
	--worker-class ${SERVER_WORKER_CLASS:-gevent} \
	--worker-connections ${SERVER_WORKER_CONNECTIONS:-10} \
	--timeout ${GUNICORN_TIMEOUT:-200} \
	app:app
	fi
	fi

[crazywoola]

cc @laipz8200

[laipz8200]

Although we've added a lock at the

dify/api/commands.py

Line 677 in a4ef900

lock = redis_client.lock(name="db_upgrade_lock", timeout=60)

to manage migration issues across multiple instances, I believe it would be beneficial to include an additional solution as well.