Обсуждение: Building Postgresql under Windows question
We are still having trouble with the service controller pg_ctl.exe vanishing after some period of time. Hence I am doing debug builds from the source tree according to the instructions found at: http://developer.postgresql.org/pgdocs/postgres/install-win32-full.html Specifically, I am using this method: It is also possible to build from inside the Visual Studio GUI. In this case, you need to run: perl mkvcbuild.pl There are a few projects that will not build (only 7 out of 99). I have not installed ossp-uuid yet, so one of the failures is expected. On the CONTRIB projects I am not concerned at all. However, on the internationalization failures, I want to know what the failures mean. Here is the output of the debug build process from the Visual Studio IDE (includes all failures): 1>------ Build started: Project: utf8_and_euc_kr, Configuration: Debug Win32 ------ 1>Generate DEF file 1>Not re-generating UTF8_AND_EUC_KR.DEF, file already exists. 1>Linking... 1>utf8_and_euc_kr.def : error LNK2001: unresolved external symbol euc_jp_to_utf8 1>utf8_and_euc_kr.def : error LNK2001: unresolved external symbol pg_finfo_euc_jp_to_utf8 1>utf8_and_euc_kr.def : error LNK2001: unresolved external symbol pg_finfo_utf8_to_euc_jp 1>utf8_and_euc_kr.def : error LNK2001: unresolved external symbol utf8_to_euc_jp 1>Debug\utf8_and_euc_kr\utf8_and_euc_kr.lib : fatal error LNK1120: 4 unresolved externals 2>------ Build started: Project: utf8_and_euc_cn, Configuration: Debug Win32 ------ 3>------ Build started: Project: euc_kr_and_mic, Configuration: Debug Win32 ------ 1>Build log was saved at "file://c:\dcorbit64\postgresql\postgresql-8.3.7\Debug\utf8_and_euc_kr\B uildLog.htm" 1>utf8_and_euc_kr - 5 error(s), 0 warning(s) 2>Generate DEF file 3>Generate DEF file 3>Not re-generating EUC_KR_AND_MIC.DEF, file already exists. 2>Not re-generating UTF8_AND_EUC_CN.DEF, file already exists. 3>Linking... 2>Linking... 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol euc_jp_to_mic 2>utf8_and_euc_cn.def : error LNK2001: unresolved external symbol koi8r_to_utf8 2>utf8_and_euc_cn.def : error LNK2001: unresolved external symbol pg_finfo_koi8r_to_utf8 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol euc_jp_to_sjis 2>utf8_and_euc_cn.def : error LNK2001: unresolved external symbol pg_finfo_utf8_to_koi8r 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol mic_to_euc_jp 2>utf8_and_euc_cn.def : error LNK2001: unresolved external symbol utf8_to_koi8r 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol mic_to_sjis 2>Debug\utf8_and_euc_cn\utf8_and_euc_cn.lib : fatal error LNK1120: 4 unresolved externals 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol pg_finfo_euc_jp_to_mic 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol pg_finfo_euc_jp_to_sjis 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol pg_finfo_mic_to_euc_jp 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol pg_finfo_mic_to_sjis 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol pg_finfo_sjis_to_euc_jp 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol pg_finfo_sjis_to_mic 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol sjis_to_euc_jp 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol sjis_to_mic 3>Debug\euc_kr_and_mic\euc_kr_and_mic.lib : fatal error LNK1120: 12 unresolved externals 2>Build log was saved at "file://c:\dcorbit64\postgresql\postgresql-8.3.7\Debug\utf8_and_euc_cn\B uildLog.htm" 2>utf8_and_euc_cn - 5 error(s), 0 warning(s) 4>------ Build started: Project: utf8_and_shift_jis_2004, Configuration: Debug Win32 ------ 4>Generate DEF file 3>Build log was saved at "file://c:\dcorbit64\postgresql\postgresql-8.3.7\Debug\euc_kr_and_mic\Bu ildLog.htm" 3>euc_kr_and_mic - 13 error(s), 0 warning(s) 4>Not re-generating UTF8_AND_SHIFT_JIS_2004.DEF, file already exists. 5>------ Build started: Project: uuid-ossp, Configuration: Debug Win32 ------ 5>Compiling... 4>Linking... 5>uuid-ossp.c 4>utf8_and_shift_jis_2004.def : error LNK2001: unresolved external symbol pg_finfo_uhc_to_utf8 4>utf8_and_shift_jis_2004.def : error LNK2001: unresolved external symbol pg_finfo_utf8_to_uhc 4>utf8_and_shift_jis_2004.def : error LNK2001: unresolved external symbol uhc_to_utf8 4>utf8_and_shift_jis_2004.def : error LNK2001: unresolved external symbol utf8_to_uhc 4>Debug\utf8_and_shift_jis_2004\utf8_and_shift_jis_2004.lib : fatal error LNK1120: 4 unresolved externals 5>.\contrib\uuid-ossp\uuid-ossp.c(27) : fatal error C1083: Cannot open include file: 'uuid.h': No such file or directory 5>Build log was saved at "file://c:\dcorbit64\postgresql\postgresql-8.3.7\Debug\uuid-ossp\BuildLo g.htm" 5>uuid-ossp - 1 error(s), 0 warning(s) 4>Build log was saved at "file://c:\dcorbit64\postgresql\postgresql-8.3.7\Debug\utf8_and_shift_ji s_2004\BuildLog.htm" 4>utf8_and_shift_jis_2004 - 5 error(s), 0 warning(s) 6>------ Build started: Project: hstore, Configuration: Debug Win32 ------ 6>Generate DEF file 6>Not re-generating HSTORE.DEF, file already exists. 7>------ Build started: Project: cube, Configuration: Debug Win32 ------ 7>Generate DEF file 6>Linking... 6>hstore.def : error LNK2001: unresolved external symbol lo_manage 6>hstore.def : error LNK2001: unresolved external symbol pg_finfo_lo_manage 6>Debug\hstore\hstore.lib : fatal error LNK1120: 2 unresolved externals 6>Build log was saved at "file://c:\dcorbit64\postgresql\postgresql-8.3.7\Debug\hstore\BuildLog.h tm" 6>hstore - 3 error(s), 0 warning(s) 7>Not re-generating CUBE.DEF, file already exists. 7>Linking... 7>cube.def : error LNK2001: unresolved external symbol chkpass_eq 7>cube.def : error LNK2001: unresolved external symbol chkpass_in 7>cube.def : error LNK2001: unresolved external symbol chkpass_ne 7>cube.def : error LNK2001: unresolved external symbol chkpass_out 7>cube.def : error LNK2001: unresolved external symbol chkpass_rout 7>cube.def : error LNK2001: unresolved external symbol pg_finfo_chkpass_eq 7>cube.def : error LNK2001: unresolved external symbol pg_finfo_chkpass_in 7>cube.def : error LNK2001: unresolved external symbol pg_finfo_chkpass_ne 7>cube.def : error LNK2001: unresolved external symbol pg_finfo_chkpass_out 7>cube.def : error LNK2001: unresolved external symbol pg_finfo_chkpass_rout 7>Debug\cube\cube.lib : fatal error LNK1120: 10 unresolved externals 7>Build log was saved at "file://c:\dcorbit64\postgresql\postgresql-8.3.7\Debug\cube\BuildLog.htm " 7>cube - 11 error(s), 0 warning(s) ========== Build: 0 succeeded, 7 failed, 92 up-to-date, 0 skipped ==========
Pg_ctl.exe is exiting with a success code from line 1946 of PG_CTL.C This is not appropriate behavior for a service unless shutdown has been requested. > -----Original Message----- > From: pgsql-hackers-owner@postgresql.org [mailto:pgsql-hackers- > owner@postgresql.org] On Behalf Of Dann Corbit > Sent: Tuesday, April 28, 2009 11:47 AM > To: pgsql-hackers@postgresql.org > Subject: [HACKERS] Building Postgresql under Windows question > > We are still having trouble with the service controller pg_ctl.exe > vanishing after some period of time. > > Hence I am doing debug builds from the source tree according to the > instructions found at: > http://developer.postgresql.org/pgdocs/postgres/install-win32-full.html > > Specifically, I am using this method: > It is also possible to build from inside the Visual Studio GUI. In this > case, you need to run: > perl mkvcbuild.pl > > There are a few projects that will not build (only 7 out of 99). > > I have not installed ossp-uuid yet, so one of the failures is expected. > On the CONTRIB projects I am not concerned at all. > However, on the internationalization failures, I want to know what the > failures mean. > Here is the output of the debug build process from the Visual Studio > IDE > (includes all failures): > > 1>------ Build started: Project: utf8_and_euc_kr, Configuration: Debug > Win32 ------ > 1>Generate DEF file > 1>Not re-generating UTF8_AND_EUC_KR.DEF, file already exists. > 1>Linking... > 1>utf8_and_euc_kr.def : error LNK2001: unresolved external symbol > euc_jp_to_utf8 > 1>utf8_and_euc_kr.def : error LNK2001: unresolved external symbol > pg_finfo_euc_jp_to_utf8 > 1>utf8_and_euc_kr.def : error LNK2001: unresolved external symbol > pg_finfo_utf8_to_euc_jp > 1>utf8_and_euc_kr.def : error LNK2001: unresolved external symbol > utf8_to_euc_jp > 1>Debug\utf8_and_euc_kr\utf8_and_euc_kr.lib : fatal error LNK1120: 4 > unresolved externals > 2>------ Build started: Project: utf8_and_euc_cn, Configuration: Debug > Win32 ------ > 3>------ Build started: Project: euc_kr_and_mic, Configuration: Debug > Win32 ------ > 1>Build log was saved at > "file://c:\dcorbit64\postgresql\postgresql- > 8.3.7\Debug\utf8_and_euc_kr\B > uildLog.htm" > 1>utf8_and_euc_kr - 5 error(s), 0 warning(s) > 2>Generate DEF file > 3>Generate DEF file > 3>Not re-generating EUC_KR_AND_MIC.DEF, file already exists. > 2>Not re-generating UTF8_AND_EUC_CN.DEF, file already exists. > 3>Linking... > 2>Linking... > 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol > euc_jp_to_mic > 2>utf8_and_euc_cn.def : error LNK2001: unresolved external symbol > koi8r_to_utf8 > 2>utf8_and_euc_cn.def : error LNK2001: unresolved external symbol > pg_finfo_koi8r_to_utf8 > 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol > euc_jp_to_sjis > 2>utf8_and_euc_cn.def : error LNK2001: unresolved external symbol > pg_finfo_utf8_to_koi8r > 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol > mic_to_euc_jp > 2>utf8_and_euc_cn.def : error LNK2001: unresolved external symbol > utf8_to_koi8r > 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol > mic_to_sjis > 2>Debug\utf8_and_euc_cn\utf8_and_euc_cn.lib : fatal error LNK1120: 4 > unresolved externals > 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol > pg_finfo_euc_jp_to_mic > 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol > pg_finfo_euc_jp_to_sjis > 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol > pg_finfo_mic_to_euc_jp > 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol > pg_finfo_mic_to_sjis > 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol > pg_finfo_sjis_to_euc_jp > 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol > pg_finfo_sjis_to_mic > 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol > sjis_to_euc_jp > 3>euc_kr_and_mic.def : error LNK2001: unresolved external symbol > sjis_to_mic > 3>Debug\euc_kr_and_mic\euc_kr_and_mic.lib : fatal error LNK1120: 12 > unresolved externals > 2>Build log was saved at > "file://c:\dcorbit64\postgresql\postgresql- > 8.3.7\Debug\utf8_and_euc_cn\B > uildLog.htm" > 2>utf8_and_euc_cn - 5 error(s), 0 warning(s) > 4>------ Build started: Project: utf8_and_shift_jis_2004, > Configuration: > Debug Win32 ------ > 4>Generate DEF file > 3>Build log was saved at > "file://c:\dcorbit64\postgresql\postgresql- > 8.3.7\Debug\euc_kr_and_mic\Bu > ildLog.htm" > 3>euc_kr_and_mic - 13 error(s), 0 warning(s) > 4>Not re-generating UTF8_AND_SHIFT_JIS_2004.DEF, file already exists. > 5>------ Build started: Project: uuid-ossp, Configuration: Debug Win32 > ------ > 5>Compiling... > 4>Linking... > 5>uuid-ossp.c > 4>utf8_and_shift_jis_2004.def : error LNK2001: unresolved external > symbol pg_finfo_uhc_to_utf8 > 4>utf8_and_shift_jis_2004.def : error LNK2001: unresolved external > symbol pg_finfo_utf8_to_uhc > 4>utf8_and_shift_jis_2004.def : error LNK2001: unresolved external > symbol uhc_to_utf8 > 4>utf8_and_shift_jis_2004.def : error LNK2001: unresolved external > symbol utf8_to_uhc > 4>Debug\utf8_and_shift_jis_2004\utf8_and_shift_jis_2004.lib : fatal > error LNK1120: 4 unresolved externals > 5>.\contrib\uuid-ossp\uuid-ossp.c(27) : fatal error C1083: Cannot open > include file: 'uuid.h': No such file or directory > 5>Build log was saved at > "file://c:\dcorbit64\postgresql\postgresql-8.3.7\Debug\uuid- > ossp\BuildLo > g.htm" > 5>uuid-ossp - 1 error(s), 0 warning(s) > 4>Build log was saved at > "file://c:\dcorbit64\postgresql\postgresql- > 8.3.7\Debug\utf8_and_shift_ji > s_2004\BuildLog.htm" > 4>utf8_and_shift_jis_2004 - 5 error(s), 0 warning(s) > 6>------ Build started: Project: hstore, Configuration: Debug Win32 > ------ > 6>Generate DEF file > 6>Not re-generating HSTORE.DEF, file already exists. > 7>------ Build started: Project: cube, Configuration: Debug Win32 ----- > - > 7>Generate DEF file > 6>Linking... > 6>hstore.def : error LNK2001: unresolved external symbol lo_manage > 6>hstore.def : error LNK2001: unresolved external symbol > pg_finfo_lo_manage > 6>Debug\hstore\hstore.lib : fatal error LNK1120: 2 unresolved externals > 6>Build log was saved at > "file://c:\dcorbit64\postgresql\postgresql- > 8.3.7\Debug\hstore\BuildLog.h > tm" > 6>hstore - 3 error(s), 0 warning(s) > 7>Not re-generating CUBE.DEF, file already exists. > 7>Linking... > 7>cube.def : error LNK2001: unresolved external symbol chkpass_eq > 7>cube.def : error LNK2001: unresolved external symbol chkpass_in > 7>cube.def : error LNK2001: unresolved external symbol chkpass_ne > 7>cube.def : error LNK2001: unresolved external symbol chkpass_out > 7>cube.def : error LNK2001: unresolved external symbol chkpass_rout > 7>cube.def : error LNK2001: unresolved external symbol > pg_finfo_chkpass_eq > 7>cube.def : error LNK2001: unresolved external symbol > pg_finfo_chkpass_in > 7>cube.def : error LNK2001: unresolved external symbol > pg_finfo_chkpass_ne > 7>cube.def : error LNK2001: unresolved external symbol > pg_finfo_chkpass_out > 7>cube.def : error LNK2001: unresolved external symbol > pg_finfo_chkpass_rout > 7>Debug\cube\cube.lib : fatal error LNK1120: 10 unresolved externals > 7>Build log was saved at > "file://c:\dcorbit64\postgresql\postgresql- > 8.3.7\Debug\cube\BuildLog.htm > " > 7>cube - 11 error(s), 0 warning(s) > ========== Build: 0 succeeded, 7 failed, 92 up-to-date, 0 skipped > ========== > > > -- > Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) > To make changes to your subscription: > http://www.postgresql.org/mailpref/pgsql-hackers
Dann Corbit wrote: > Pg_ctl.exe is exiting with a success code from line 1946 of PG_CTL.C > > This is not appropriate behavior for a service unless shutdown has been > requested. > pg_ctl calls *StartServiceCtrlDispatcher*(). It can only get to the line you mention when called as a service after that call returns. MSDN states: "If *StartServiceCtrlDispatcher* succeeds, it connects the calling thread to the service control manager and does not return until all running services in the process have entered the SERVICE_STOPPED state." So it appears that something is causing your service to enter that state. cheers andrew
> -----Original Message-----
> From: Andrew Dunstan [mailto:andrew@dunslane.net]
> Sent: Tuesday, April 28, 2009 12:42 PM
> To: Dann Corbit
> Cc: pgsql-hackers@postgresql.org
> Subject: Re: [HACKERS] Building Postgresql under Windows question
>
>
>
> Dann Corbit wrote:
> > Pg_ctl.exe is exiting with a success code from line 1946 of PG_CTL.C
> >
> > This is not appropriate behavior for a service unless shutdown has
> been
> > requested.
> >
>
> pg_ctl calls *StartServiceCtrlDispatcher*(). It can only get to the
> line
> you mention when called as a service after that call returns. MSDN
> states:
>
> "If *StartServiceCtrlDispatcher* succeeds, it connects the calling
> thread to the service control manager and does not return until all
> running services in the process have entered the SERVICE_STOPPED
> state."
>
> So it appears that something is causing your service to enter that
> state.
It is interesting that it happens even if I run no queries at all.
This is the only reference to the service control dispatcher I can find
in pg_ctl.c:
static void
pgwin32_doRunAsService(void)
{SERVICE_TABLE_ENTRY st[] = {{register_servicename,
pgwin32_ServiceMain},{NULL, NULL}};
if (StartServiceCtrlDispatcher(st) == 0){ write_stderr(_("%s: could not start service \"%s\":
error code %d\n"), progname, register_servicename, (int)
GetLastError()); exit(1);}
}
BTW, the exit(1) calls should be exit(EXIT_FAILURE) though there will be
no difficulty on any POSIX system.
> -----Original Message-----
> From: Andrew Dunstan [mailto:andrew@dunslane.net]
> Sent: Tuesday, April 28, 2009 12:42 PM
> To: Dann Corbit
> Cc: pgsql-hackers@postgresql.org
> Subject: Re: [HACKERS] Building Postgresql under Windows question
>
>
>
> Dann Corbit wrote:
> > Pg_ctl.exe is exiting with a success code from line 1946 of PG_CTL.C
> >
> > This is not appropriate behavior for a service unless shutdown has
> been
> > requested.
> >
>
> pg_ctl calls *StartServiceCtrlDispatcher*(). It can only get to the
> line
> you mention when called as a service after that call returns. MSDN
> states:
>
> "If *StartServiceCtrlDispatcher* succeeds, it connects the calling
> thread to the service control manager and does not return until all
> running services in the process have entered the SERVICE_STOPPED
> state."
>
> So it appears that something is causing your service to enter that
> state.
It appears that SERVICE_STOPPED comes from here:
static void WINAPI
pgwin32_ServiceMain(DWORD argc, LPTSTR * argv)
{PROCESS_INFORMATION pi;DWORD ret;DWORD check_point_start;
/* Initialize variables */status.dwWin32ExitCode = S_OK;status.dwCheckPoint = 0;status.dwWaitHint =
60000;status.dwServiceType= SERVICE_WIN32_OWN_PROCESS;status.dwControlsAccepted = SERVICE_ACCEPT_STOP |
SERVICE_ACCEPT_SHUTDOWN | SERVICE_ACCEPT_PAUSE_CONTINUE;status.dwServiceSpecificExitCode = 0;status.dwCurrentState =
SERVICE_START_PENDING;
memset(&pi, 0, sizeof(pi));
read_post_opts();
/* Register the control request handler */if ((hStatus = RegisterServiceCtrlHandler(register_servicename,
pgwin32_ServiceHandler)) == (SERVICE_STATUS_HANDLE) 0) return;
if ((shutdownEvent = CreateEvent(NULL, true, false, NULL)) ==
NULL) return;
/* Start the postmaster */pgwin32_SetServiceStatus(SERVICE_START_PENDING);if
(!CreateRestrictedProcess(pgwin32_CommandLine(false),&pi,
true)){ pgwin32_SetServiceStatus(SERVICE_STOPPED); return;}postmasterPID = pi.dwProcessId;postmasterProcess =
pi.hProcess;CloseHandle(pi.hThread);
if (do_wait){ write_eventlog(EVENTLOG_INFORMATION_TYPE, _("Waiting for
server startup...\n")); if (test_postmaster_connection(true) == false) {
write_eventlog(EVENTLOG_INFORMATION_TYPE,
_("Timed out waiting for server startup\n")); pgwin32_SetServiceStatus(SERVICE_STOPPED); return; }
write_eventlog(EVENTLOG_INFORMATION_TYPE,_("Server
started and accepting connections\n"));}
/* * Save the checkpoint value as it might have been incremented
in * test_postmaster_connection */check_point_start = status.dwCheckPoint;
pgwin32_SetServiceStatus(SERVICE_RUNNING);
/* Wait for quit... */ret = WaitForMultipleObjects(2, shutdownHandles, FALSE,
INFINITE);
pgwin32_SetServiceStatus(SERVICE_STOP_PENDING);switch (ret){ case WAIT_OBJECT_0: /* shutdown event */
kill(postmasterPID,SIGINT);
/* * Increment the checkpoint and try again Abort
after 12 * checkpoints as the postmaster has probably
hung */ while (WaitForSingleObject(postmasterProcess,
5000) == WAIT_TIMEOUT && status.dwCheckPoint < 12) status.dwCheckPoint++; break;
case (WAIT_OBJECT_0 + 1): /* postmaster
went down */ break;
default: /* shouldn't get here? */ break;}
CloseHandle(shutdownEvent);CloseHandle(postmasterProcess);
pgwin32_SetServiceStatus(SERVICE_STOPPED);
}
I will set a breakpoint on every place that the status is set to
SERVICE_STOPPED and report what I have found.
It's this one:
if (do_wait)
{
write_eventlog(EVENTLOG_INFORMATION_TYPE, _("Waiting for server
startup...\n"));
if (test_postmaster_connection(true) == false)
{
write_eventlog(EVENTLOG_INFORMATION_TYPE, _("Timed out
waiting for server startup\n"));
pgwin32_SetServiceStatus(SERVICE_STOPPED); //
<<<<<<<<<<<<<<<<<<<<<<<<<<< HERE!
return;
}
write_eventlog(EVENTLOG_INFORMATION_TYPE, _("Server started and
accepting connections\n"));
}
The attached bitmap shows that the servers are running, though
pg_ctl.exe has exited.
I am also able to attach and query.
> -----Original Message-----
> From: pgsql-hackers-owner@postgresql.org [mailto:pgsql-hackers-
> owner@postgresql.org] On Behalf Of Dann Corbit
> Sent: Tuesday, April 28, 2009 12:49 PM
> To: Andrew Dunstan
> Cc: pgsql-hackers@postgresql.org
> Subject: Re: [HACKERS] Building Postgresql under Windows question
>
> > -----Original Message-----
> > From: Andrew Dunstan [mailto:andrew@dunslane.net]
> > Sent: Tuesday, April 28, 2009 12:42 PM
> > To: Dann Corbit
> > Cc: pgsql-hackers@postgresql.org
> > Subject: Re: [HACKERS] Building Postgresql under Windows question
> >
> >
> >
> > Dann Corbit wrote:
> > > Pg_ctl.exe is exiting with a success code from line 1946 of
> PG_CTL.C
> > >
> > > This is not appropriate behavior for a service unless shutdown has
> > been
> > > requested.
> > >
> >
> > pg_ctl calls *StartServiceCtrlDispatcher*(). It can only get to the
> > line
> > you mention when called as a service after that call returns. MSDN
> > states:
> >
> > "If *StartServiceCtrlDispatcher* succeeds, it connects the calling
> > thread to the service control manager and does not return until all
> > running services in the process have entered the SERVICE_STOPPED
> > state."
> >
> > So it appears that something is causing your service to enter that
> > state.
>
> It appears that SERVICE_STOPPED comes from here:
>
> static void WINAPI
> pgwin32_ServiceMain(DWORD argc, LPTSTR * argv)
> {
> PROCESS_INFORMATION pi;
> DWORD ret;
> DWORD check_point_start;
>
> /* Initialize variables */
> status.dwWin32ExitCode = S_OK;
> status.dwCheckPoint = 0;
> status.dwWaitHint = 60000;
> status.dwServiceType = SERVICE_WIN32_OWN_PROCESS;
> status.dwControlsAccepted = SERVICE_ACCEPT_STOP |
> SERVICE_ACCEPT_SHUTDOWN | SERVICE_ACCEPT_PAUSE_CONTINUE;
> status.dwServiceSpecificExitCode = 0;
> status.dwCurrentState = SERVICE_START_PENDING;
>
> memset(&pi, 0, sizeof(pi));
>
> read_post_opts();
>
> /* Register the control request handler */
> if ((hStatus = RegisterServiceCtrlHandler(register_servicename,
> pgwin32_ServiceHandler)) == (SERVICE_STATUS_HANDLE) 0)
> return;
>
> if ((shutdownEvent = CreateEvent(NULL, true, false, NULL)) ==
> NULL)
> return;
>
> /* Start the postmaster */
> pgwin32_SetServiceStatus(SERVICE_START_PENDING);
> if (!CreateRestrictedProcess(pgwin32_CommandLine(false), &pi,
> true))
> {
> pgwin32_SetServiceStatus(SERVICE_STOPPED);
> return;
> }
> postmasterPID = pi.dwProcessId;
> postmasterProcess = pi.hProcess;
> CloseHandle(pi.hThread);
>
> if (do_wait)
> {
> write_eventlog(EVENTLOG_INFORMATION_TYPE, _("Waiting for
> server startup...\n"));
> if (test_postmaster_connection(true) == false)
> {
> write_eventlog(EVENTLOG_INFORMATION_TYPE,
> _("Timed out waiting for server startup\n"));
> pgwin32_SetServiceStatus(SERVICE_STOPPED);
> return;
> }
> write_eventlog(EVENTLOG_INFORMATION_TYPE, _("Server
> started and accepting connections\n"));
> }
>
> /*
> * Save the checkpoint value as it might have been incremented
> in
> * test_postmaster_connection
> */
> check_point_start = status.dwCheckPoint;
>
> pgwin32_SetServiceStatus(SERVICE_RUNNING);
>
> /* Wait for quit... */
> ret = WaitForMultipleObjects(2, shutdownHandles, FALSE,
> INFINITE);
>
> pgwin32_SetServiceStatus(SERVICE_STOP_PENDING);
> switch (ret)
> {
> case WAIT_OBJECT_0: /* shutdown event */
> kill(postmasterPID, SIGINT);
>
> /*
> * Increment the checkpoint and try again Abort
> after 12
> * checkpoints as the postmaster has probably
> hung
> */
> while (WaitForSingleObject(postmasterProcess,
> 5000) == WAIT_TIMEOUT && status.dwCheckPoint < 12)
> status.dwCheckPoint++;
> break;
>
> case (WAIT_OBJECT_0 + 1): /* postmaster
> went down */
> break;
>
> default:
> /* shouldn't get here? */
> break;
> }
>
> CloseHandle(shutdownEvent);
> CloseHandle(postmasterProcess);
>
> pgwin32_SetServiceStatus(SERVICE_STOPPED);
> }
>
> I will set a breakpoint on every place that the status is set to
> SERVICE_STOPPED and report what I have found.
>
>
> --
> Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-hackers
Вложения
In this function:
static bool
test_postmaster_connection(bool do_checkpoint)
This code will never succeed:snprintf(connstr, sizeof(connstr), "dbname=postgres port=%s connect_timeout=5",
portstr);
for (i = 0; i < wait_seconds; i++){ if ((conn = PQconnectdb(connstr)) != NULL && (PQstatus(conn) ==
CONNECTION_OK|| PQconnectionNeedsPassword(conn))) { PQfinish(conn); success = true;
break; }
Because pg_hba.conf has this:
# TYPE DATABASE USER CIDR-ADDRESS METHOD
# IPv4 local connections:
host all all 127.0.0.1/32 trust
# IPv6 local connections:
host all all ::1/128 trust
to allow connections from the local machine.
(A password is not needed to connect, so the code always fails).
Dann Corbit wrote:
> In this function:
> static bool
> test_postmaster_connection(bool do_checkpoint)
>
> This code will never succeed:
> snprintf(connstr, sizeof(connstr),
> "dbname=postgres port=%s connect_timeout=5",
> portstr);
>
> for (i = 0; i < wait_seconds; i++)
> {
> if ((conn = PQconnectdb(connstr)) != NULL &&
> (PQstatus(conn) == CONNECTION_OK ||
> PQconnectionNeedsPassword(conn)))
> {
> PQfinish(conn);
> success = true;
> break;
> }
>
>
> Because pg_hba.conf has this:
> # TYPE DATABASE USER CIDR-ADDRESS METHOD
>
> # IPv4 local connections:
> host all all 127.0.0.1/32 trust
> # IPv6 local connections:
> host all all ::1/128 trust
>
> to allow connections from the local machine.
>
> (A password is not needed to connect, so the code always fails).
>
>
Why won't
PQstatus(conn) == CONNECTION_OK
be true and thus the code will succeed without requiring a password?
cheers
andrew
> -----Original Message-----
> From: Andrew Dunstan [mailto:andrew@dunslane.net]
> Sent: Tuesday, April 28, 2009 2:27 PM
> To: Dann Corbit
> Cc: pgsql-hackers@postgresql.org
> Subject: Re: [HACKERS] Building Postgresql under Windows question
>
>
>
> Dann Corbit wrote:
> > In this function:
> > static bool
> > test_postmaster_connection(bool do_checkpoint)
> >
> > This code will never succeed:
> > snprintf(connstr, sizeof(connstr),
> > "dbname=postgres port=%s connect_timeout=5",
> > portstr);
> >
> > for (i = 0; i < wait_seconds; i++)
> > {
> > if ((conn = PQconnectdb(connstr)) != NULL &&
> > (PQstatus(conn) == CONNECTION_OK ||
> > PQconnectionNeedsPassword(conn)))
> > {
> > PQfinish(conn);
> > success = true;
> > break;
> > }
> >
> >
> > Because pg_hba.conf has this:
> > # TYPE DATABASE USER CIDR-ADDRESS METHOD
> >
> > # IPv4 local connections:
> > host all all 127.0.0.1/32 trust
> > # IPv6 local connections:
> > host all all ::1/128 trust
> >
> > to allow connections from the local machine.
> >
> > (A password is not needed to connect, so the code always fails).
> >
> >
>
>
> Why won't
>
> PQstatus(conn) == CONNECTION_OK
>
> be true and thus the code will succeed without requiring a password?
It returns the value CONNECTION_STARTED
I have found a work-around for now.
If I set the method to password in pg_hba.conf, the service starts and
runs correctly.
It is only when the method is set to trust that we get 100% failures.
Dann Corbit wrote:
>> -----Original Message-----
>> From: Andrew Dunstan [mailto:andrew@dunslane.net]
>> Sent: Tuesday, April 28, 2009 2:27 PM
>> To: Dann Corbit
>> Cc: pgsql-hackers@postgresql.org
>> Subject: Re: [HACKERS] Building Postgresql under Windows question
>>
>>
>>
>> Dann Corbit wrote:
>>
>>> In this function:
>>> static bool
>>> test_postmaster_connection(bool do_checkpoint)
>>>
>>> This code will never succeed:
>>> snprintf(connstr, sizeof(connstr),
>>> "dbname=postgres port=%s connect_timeout=5",
>>> portstr);
>>>
>>> for (i = 0; i < wait_seconds; i++)
>>> {
>>> if ((conn = PQconnectdb(connstr)) != NULL &&
>>> (PQstatus(conn) == CONNECTION_OK ||
>>> PQconnectionNeedsPassword(conn)))
>>> {
>>> PQfinish(conn);
>>> success = true;
>>> break;
>>> }
>>>
>>>
>>> Because pg_hba.conf has this:
>>> # TYPE DATABASE USER CIDR-ADDRESS METHOD
>>>
>>> # IPv4 local connections:
>>> host all all 127.0.0.1/32 trust
>>> # IPv6 local connections:
>>> host all all ::1/128 trust
>>>
>>> to allow connections from the local machine.
>>>
>>> (A password is not needed to connect, so the code always fails).
>>>
>>>
>>>
>> Why won't
>>
>> PQstatus(conn) == CONNECTION_OK
>>
>> be true and thus the code will succeed without requiring a password?
>>
>
> It returns the value CONNECTION_STARTED
>
> I have found a work-around for now.
>
> If I set the method to password in pg_hba.conf, the service starts and
> runs correctly.
>
> It is only when the method is set to trust that we get 100% failures.
>
>
>
Then that looks like a libpq bug :-( According to the docs:
"The status can be one of a number of values. However, only two of these
are seen outside of an asynchronous connection procedure: CONNECTION_OK
and CONNECTION_BAD."
cheers
andrew
"Dann Corbit" <DCorbit@connx.com> writes:
>> From: Andrew Dunstan [mailto:andrew@dunslane.net]
>> Why won't
>> PQstatus(conn) == CONNECTION_OK
>> be true and thus the code will succeed without requiring a password?
> It returns the value CONNECTION_STARTED
It certainly shouldn't. You're effectively asserting that PQconnectdb
is broken for everyone on every platform, which is demonstrably not the
case. Are you fooling with modified libpq code by any chance?
regards, tom lane