The following bug has been logged online:
Bug reference: 5892
Logged by: Harsh Sinha
Email address: harsh.sinha@hp.com
PostgreSQL version: 8.3.7
Operating system: Windows
Description: Restart Failed after WAL restore
Details:
Hi,
We have implemented Continuous Archiving using the WAL technique. We have
tested it to be working. There is one condition where we are trying to
restore with around 34 WAL files. After the file transfers and creation of
the recovery.conf file. We start the PostgreSQL service from the Windows
services.msc
We also see the message "Archive Recovery Complete" in the logs under pg_log
folder. But Since restore takes a long time, The service manager reports
that the service has stopped, when actually after restore its running. Now
when we start the service again we get error:
Lock file 'postmaster.pid' already exists....
My analysis is that since windows gave up starting postgres, it put the
status as STOPPED. When restore completed after a long time Postgres created
postmaster.pid and started. Now when we try to start windows sees that there
is a postmaster file. This creates a deadlock situation.
Please let me know if you are aware of this issue or can help me solve this
prob
