On Jul 8 my host server encountered an unexpected power event that caused the server to restart abruptly. (Somehow the event made it past my UPS, but that's a different discussion.)

Here are the logs from that date, starting just after the unexpected restart:

Upon that restart, a running phantom PostgreSQL instance prevented the main BDRSuite service from starting (due to the PostgreSQL service failure). I didn't discover all of this until today, after which I killed the phantom process and started all of the other services (at which point the missed scheduled backup commenced and completed successfully). All of them, however, save the GranularRecovery service, which now won't start.

There are no corresponding entries in the Windows Application Log, but there are some concerning entries in the PostgreSQL log. Apparently there's some sort of schema problem, which is odd since I wasn't performing any sort of upgrade/update when the power event occurred.

In case it helps, here're the log files I have:

(Note that the timestamps are in Alaska Daylight Time, currently -8:00.)

An additional hardware detail: the server's storage controller is a Dell PERC H200, which doesn't have a battery-backed cache. So it's possible that an uncommitted write was lost during the power event (although a backup wasn't running at that time).

I'm attaching today's logs (postgresql-2026-07-12_201041.log) up to the reboot I performed about a couple of hours ago.

What might I do to get this service to start?

Thanks.