Fix race condition in job recovery causing external_id lookup failure

[mvdbeek]

The ManagerMonitor thread was starting before active jobs had their external IDs recovered from disk. This caused status checks to fail with "Failed to obtain external_id for job_id" errors on startup.

Observed on usegalaxy.be with job 655740:

• Monitor thread checked job status at 17:10:56,273
• MainThread recovered external_id (292) at 17:10:56,278
• The 5ms gap caused the status check to fail

2026-02-26 17:10:56,273 ERROR [pulsar.managers.stateful][[manager=production]-[action=monitor]] Active job IDs in directory '/data/share/persisted_data/production-active-jobs': ['655740']
2026-02-26 17:10:56,273 ERROR [pulsar.managers.stateful][[manager=production]-[action=monitor]] Failed checking active job status for job_id 655740
Traceback (most recent call last):
  File "/opt/pulsar/venv3/lib64/python3.9/site-packages/pulsar/managers/stateful.py", line 383, in _monitor_active_jobs
    self._check_active_job_status(active_job_id)
  File "/opt/pulsar/venv3/lib64/python3.9/site-packages/pulsar/managers/stateful.py", line 397, in _check_active_job_status
    self.stateful_manager.get_status(active_job_id)
  File "/opt/pulsar/venv3/lib64/python3.9/site-packages/pulsar/managers/stateful.py", line 165, in get_status
    proxy_status, state_change = self.__proxy_status(job_directory, job_id)
  File "/opt/pulsar/venv3/lib64/python3.9/site-packages/pulsar/managers/stateful.py", line 189, in __proxy_status
    proxy_status = self._proxied_manager.get_status(job_id)
  File "/opt/pulsar/venv3/lib64/python3.9/site-packages/pulsar/managers/queued_condor.py", line 68, in get_status
    raise Exception("Failed to obtain external_id for job_id %s, cannot determine status." % job_id)
Exception: Failed to obtain external_id for job_id 655740, cannot determine status.
2026-02-26 17:10:56,274 DEBUG [pulsar.client.amqp_exchange][consume-status-pyamqp://galaxy_vib:********@usegalaxy.be:5671//pulsar/galaxy_vib?ssl=1] Consuming queue '<unbound Queue pulsar_test__status -> <unbound Exchange pulsar(direct)> -> pulsar_test__status>'
2026-02-26 17:10:56,273 DEBUG [pulsar.client.amqp_exchange][consume-kill-pyamqp://galaxy_vib:********@usegalaxy.be:5671//pulsar/galaxy_vib?ssl=1] Consuming queue '<unbound Queue pulsar_test__kill -> <unbound Exchange pulsar(direct)> -> pulsar_test__kill>'
2026-02-26 17:10:56,276 ERROR [pulsar.managers.stateful][MainThread] Active job IDs in directory '/data/share/persisted_data/benchmarking-preprocessing-jobs': []
2026-02-26 17:10:56,276 ERROR [pulsar.managers.stateful][[manager=test]-[action=monitor]] Active job IDs in directory '/data/share/persisted_data/test-active-jobs': []
2026-02-26 17:10:56,276 ERROR [pulsar.managers.stateful][MainThread] Active job IDs in directory '/data/share/persisted_data/benchmarking-active-jobs': []
2026-02-26 17:10:56,277 ERROR [pulsar.managers.stateful][MainThread] Active job IDs in directory '/data/share/persisted_data/production-preprocessing-jobs': []
2026-02-26 17:10:56,277 ERROR [pulsar.managers.stateful][MainThread] Active job IDs in directory '/data/share/persisted_data/production-active-jobs': ['655740']
2026-02-26 17:10:56,278 ERROR [pulsar.managers.base.external][MainThread] Recovered external ID for job_id [655740]: 292
2026-02-26 17:10:56,278 ERROR [pulsar.managers.stateful][MainThread] Active job IDs in directory '/data/share/persisted_data/test-preprocessing-jobs': []
2026-02-26 17:10:56,278 ERROR [pulsar.managers.stateful][MainThread] Active job IDs in directory '/data/share/persisted_data/test-active-jobs': []

Fix: Reorder startup so __recover_jobs() runs before __setup_bind_to_message_queue(), ensuring external IDs are loaded before the monitor thread begins polling.

This is probably just a cosmetic / correctness fix since the thread would retry later.