I tried again and got the same error "Connection attempt timed out.". Log file is empty until I restart service, than I get the following. This is the complete logfile.
2019-03-12 13:41:41.246 EDT [5948] LOG: received fast shutdown request
2019-03-12 13:41:41.287 EDT [5948] LOG: aborting any active transactions
2019-03-12 13:41:41.295 EDT [5948] LOG: background worker "logical replication launcher" (PID 19912) exited with exit code 1
2019-03-12 13:41:41.336 EDT [17468] LOG: shutting down
2019-03-12 13:41:41.533 EDT [5948] LOG: database system is shut down
Since I've never heard of this strange behaviour among my
customers (we have some hundred postgres 10.x windows 7/8/10
installations), I'll try to guess and hope to find some
interesting element to investigate.
Some points:
1. When did you start the server? it's missing of the initial,
server startup, part... something stating at least that the system
is starting, even restarting from a not-proper shutdown and, at
last, "the database system is ready to accept connections". I
would say that the first part should be taken from yesterday's
file (or the file from the day before the one the file's
referencing), as if you started server yesterday (or back in the
past) and stopped today at 13.41 EDT. The log line I set in bold
states that "something" has requested the system to shut down
(Windows restart, service stop, even antivirus system (but that
has also eaten your bins, so it won't restart), something else),
but if you just started the server it can't be the first
line in log file.
2. From where are you connecting? From a remote (not localhost)
client or the server itself?
3. Did you alter config files? Default listening port is 5432,
but if you install more than one cluster you'll have to change
them.
3. Did you take a look at the Event Viewer? Is there any error?
4. Have you thought to reinstall?
Cheers
Moreno.-
That's the intended behavior on windowsThen a new log is opened when service restarts.
Thanks,
Michael Cassell
On March 12, 2019 at 12:40 PM Moreno Andreo <moreno.andreo@evolu-s.it> wrote:
Il 08/03/2019 14:28, mcassell@trimind.com ha scritto:
I have installed PostgreSQL on two different Windows 10 systems and am having the same problem. The server work great when I first connect to it but if I disconnect client and come back a few hours or a day later I am unable to connect to the service with a timeout error. I have used both pgAdmin and DBeaver as clients.
If I do a restart on the window service everything works fine again until I disconnect the client for a extended period. Can't find anywhere else this issue is discussed.
Have used both the latest 11.2 release and an older 10.x release and both operate the same. Is this a known issue? Is there a solution?
Have you checked logs (Event viewer or the folder /data/pg_log)?
Cheers
Moreno.-