Postgres Crashes and Leaves PT3; AHK unstable.

[stillnaive]

I'm currently running postgres v.8.2.5 -- the version that was distributed with the second beta of PT3 -- and when my sessions reach about 1000 hands, postgres crashes and it sends my system into a spiral of instability. of course, this leaves me to muck pocket aces and sets and stuff because i run good except when my computer is crashing i have attached the error logs here, and i'm hoping someone can help me interpret what, exactly, is going on. do i need to reinstall windows? am i just screwed here?

fwiw, my system is 1hr stable on OCCT's mem test and cpu stability test. It's not overclocked, and memory usage was not at 100% when it crashed.

one error log says the following:

[quoteppr]2008-03-02 12:27:24 STATEMENT: COPY holdem_hand_summary FROM STDIN;
2008-03-02 12:27:30 WARNING: there is no transaction in progress
2008-03-02 12:33:46 LOG: server process (PID 3112) exited with exit code -1073741502
2008-03-02 12:33:46 LOG: terminating any other active server processes
2008-03-02 12:33:46 WARNING: terminating connection because of crash of another server process
2008-03-02 12:33:46 DETAIL: The postmaster has commanded this server process to roll back the current transaction and exit, because another server process exited abnormally and possibly corrupted shared memory.
2008-03-02 12:33:46 HINT: In a moment you should be able to reconnect to the database and repeat your command.
2008-03-02 12:33:46 WARNING: terminating connection because of crash of another server process
2008-03-02 12:33:46 DETAIL: The postmaster has commanded this server process to roll back the current transaction and exit, because another server process exited abnormally and possibly corrupted shared memory.
2008-03-02 12:33:46 HINT: In a moment you should be able to reconnect to the database and repeat your command.
2008-03-02 12:33:46 WARNING: terminating connection because of crash of another server process
2008-03-02 12:33:46 DETAIL: The postmaster has commanded this server process to roll back the current transaction and exit, because another server process exited abnormally and possibly corrupted shared memory.
2008-03-02 12:33:46 HINT: In a moment you should be able to reconnect to the database and repeat your command.
2008-03-02 12:33:46 WARNING: terminating connection because of crash of another server process
2008-03-02 12:33:46 DETAIL: The postmaster has commanded this server process to roll back the current transaction and exit, because another server process exited abnormally and possibly corrupted shared memory.
2008-03-02 12:33:46 HINT: In a moment you should be able to reconnect to the database and repeat your command.
2008-03-02 12:33:46 WARNING: terminating connection because of crash of another server process
2008-03-02 12:33:46 DETAIL: The postmaster has commanded this server process to roll back the current transaction and exit, because another server process exited abnormally and possibly corrupted shared memory.
2008-03-02 12:33:46 HINT: In a moment you should be able to reconnect to the database and repeat your command.
2008-03-02 12:33:46 LOG: all server processes terminated; reinitializing
2008-03-02 12:33:46 LOG: unrecognized win32 error code: 1455
2008-03-02 12:33:46 FATAL: shmat(id=1908) failed: Invalid argument
2008-03-02 12:33:47 LOG: logger shutting down[/quoteppr]

Another said this:
[quoteppr]2008-03-02 15:00:51 LOG: CreateProcess call failed: A blocking operation was interrupted by a call to WSACancelBlockingCall.
(error code 1450)
2008-03-02 15:00:51 LOG: could not fork new process for connection: A blocking operation was interrupted by a call to WSACancelBlockingCall.
2008-03-02 15:00:51 LOG: CreateProcess call failed: A blocking operation was interrupted by a call to WSACancelBlockingCall.
(error code 1450)
2008-03-02 15:00:51 LOG: could not fork new process for connection: A blocking operation was interrupted by a call to WSACancelBlockingCall.
2008-03-02 15:00:52 LOG: could not open temporary statistics file "global/pgstat.tmp": Invalid argument
2008-03-02 15:00:58 LOG: could not open temporary statistics file "global/pgstat.tmp": Invalid argument
2008-03-02 15:01:00 LOG: could not open temporary statistics file "global/pgstat.tmp": Invalid argument
2008-03-02 15:01:05 LOG: could not open temporary statistics file "global/pgstat.tmp": Invalid argument
2008-03-02 15:01:51 LOG: CreateProcess call failed: A blocking operation was interrupted by a call to WSACancelBlockingCall.
(error code 1450)
2008-03-02 15:01:51 LOG: could not fork autovacuum process: A blocking operation was interrupted by a call to WSACancelBlockingCall.
2008-03-02 15:02:44 LOG: CreateProcess call failed: A blocking operation was interrupted by a call to WSACancelBlockingCall.
(error code 1450)
2008-03-02 15:02:44 LOG: could not fork new process for connection: A blocking operation was interrupted by a call to WSACancelBlockingCall.
2008-03-02 15:02:44 LOG: CreateProcess call failed: A blocking operation was interrupted by a call to WSACancelBlockingCall.
(error code 1450)
2008-03-02 15:02:44 LOG: could not fork new process for connection: A blocking operation was interrupted by a call to WSACancelBlockingCall.
2008-03-02 15:03:44 LOG: CreateProcess call failed: A blocking operation was interrupted by a call to WSACancelBlockingCall.
(error code 1450)
2008-03-02 15:03:44 LOG: could not fork autovacuum process: A blocking operation was interrupted by a call to WSACancelBlockingCall.[/quoteppr]

Thanks in advance!

[APerfect10]

Are you the same use who posted this on 2p2?

-Derek

[stillnaive]

yeah, i figured i'd put it up on this board because it was kinda getting caught up in the thread. since this forum deals directly with postgres, it seemed applicable!

[APerfect10]

And this happens frequently? Has it ever happened while not using AHK HUD?

-Derek

[stillnaive]

it's kind of a catch 22; i'm not importing hands or running sessions without a hud, and i can't rule it out without the hud. i'm typically not going to 9 table without being able to see any stats. the problem is that neither memory nor processor were maxed out when it happened; i don't know that it's a system resources problem, but i can't rule anything out at this point.

and yes, it happens frequently

[APerfect10]

I understand the predicament you are in, unfortunately we need to pinpoint which program is causing the problem before we can begin to troubleshoot the issue (if need be).

Would it be possible for you to import your hands into PT2 & PT3 at the same time and use PAHud with PT2 for a session and see if PT3 crashes?

Best regards,

Derek

[_dave_]

On which site do you play?

EDIT: And yes, this is the best to post to get this issue solved - it is too complex for the big threads on 2p2, needs a thread of it's own - and while it is mostly the same people here and there, it is easier to keep track of things in a single specific thread.

[APerfect10]

[quote="_dave_"jne]On which site do you play?

EDIT: And yes, this is the best to post to get this issue solved - it is too complex for the big threads on 2p2, needs a thread of it's own - and while it is mostly the same people here and there, it is easier to keep track of things in a single specific thread.[/quotejne]

PT3 Forums >>> 2p2 Forums

[_dave_]

heh, depends what for The threaded PM system here is the best I've seen imo.

Anyhow, I was thinking a suitable test for OP may be to fire up 12 FTP or Party(+FPHG) tables, leave that auto-importing for an hour or two while he isn't playing... that would eliminate AHK from the situation - but as I said in the other thread it looks to me like it may be maxing out it's postgres connections - but that shouldn't cause a fatal for postgres, it should just deny them.

[_dave_]

Another random thought - something I do not know... Is there a limit to the number of processes that may run on a Win32 box? AHK-HUD can certainly create a good few processes, as can postgres...