Movatterモバイル変換

[0]ホーム
URL:

Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

 Sign up
Appearance settings

Commita151a5c

Browse files
committed
Fix "pg_ctl start -w" to test child process status directly.
pg_ctl start with -w previously relied on a heuristic that the postmasterwould surely always manage to create postmaster.pid within five seconds.Unfortunately, that fails much more often than we would like on some of theslower, more heavily loaded buildfarm members.We have known for quite some time that we could remove the need for thatheuristic on Unix by using fork/exec instead of system() to launch thepostmaster. This allows us to know the exact PID of the postmaster, whichallows near-certain verification that the postmaster.pid file is the onewe want and not a leftover, and it also lets us use waitpid() to detectreliably whether the child postmaster has exited or not.What was blocking this change was not wanting to rewrite the Windowsversion of start_postmaster() to avoid use of CMD.EXE. That's doablein theory but would require fooling about with stdout/stderr redirection,and getting the handling of quote-containing postmaster switches tostay the same might be rather ticklish. However, we realized thatwe don't have to do that to fix the problem, because we can testwhether the shell process has exited as a proxy for whether thepostmaster is still alive. That doesn't allow an exact check of thePID in postmaster.pid, but we're no worse off than before in thatrespect; and we do get to get rid of the heuristic about how long thepostmaster might take to create postmaster.pid.On Unix, this change means that a second "pg_ctl start -w" immediatelyafter another such command will now reliably fail, whereas previouslyit would succeed if done within two seconds of the earlier command.Since that's a saner behavior anyway, it's fine. On Windows, the case canstill succeed within the same time window, since pg_ctl can't tell that theearlier postmaster's postmaster.pid isn't the pidfile it is looking for.To ensure stable test results on Windows, we can insert a short sleep intothe test script for pg_ctl, ensuring that the existing pidfile looks stale.This hack can be removed if we ever do rewrite start_postmaster(), but thatno longer seems like a high-priority thing to do.Back-patch to all supported versions, both because the current behavioris buggy and because we must do that if we want the buildfarm failuresto go away.Tom Lane and Michael Paquier
1 parentf75c4fc commita151a5c

File tree

2 files changed

+114
-85
lines changed

2 files changed

+114
-85
lines changed

‎src/bin/pg_ctl/pg_ctl.c

Lines changed: 107 additions & 83 deletions
Original file line numberDiff line numberDiff line change
@@ -28,6 +28,7 @@
2828
#include<time.h>
2929
#include<sys/types.h>
3030
#include<sys/stat.h>
31+
#include<sys/wait.h>
3132
#include<unistd.h>
3233

3334
#ifdefHAVE_SYS_RESOURCE_H
@@ -153,10 +154,10 @@ static intCreateRestrictedProcess(char *cmd, PROCESS_INFORMATION *processInfo,
153154
staticpgpid_tget_pgpid(boolis_status_request);
154155
staticchar**readfile(constchar*path);
155156
staticvoidfree_readfile(char**optlines);
156-
staticintstart_postmaster(void);
157+
staticpgpid_tstart_postmaster(void);
157158
staticvoidread_post_opts(void);
158159

159-
staticPGPingtest_postmaster_connection(bool);
160+
staticPGPingtest_postmaster_connection(pgpid_tpm_pid,booldo_checkpoint);
160161
staticboolpostmaster_is_alive(pid_tpid);
161162

162163
#if defined(HAVE_GETRLIMIT)&& defined(RLIMIT_CORE)
@@ -419,36 +420,73 @@ free_readfile(char **optlines)
419420
* start/test/stop routines
420421
*/
421422

422-
staticint
423+
/*
424+
* Start the postmaster and return its PID.
425+
*
426+
* Currently, on Windows what we return is the PID of the shell process
427+
* that launched the postmaster (and, we trust, is waiting for it to exit).
428+
* So the PID is usable for "is the postmaster still running" checks,
429+
* but cannot be compared directly to postmaster.pid.
430+
*
431+
* On Windows, we also save aside a handle to the shell process in
432+
* "postmasterProcess", which the caller should close when done with it.
433+
*/
434+
staticpgpid_t
423435
start_postmaster(void)
424436
{
425437
charcmd[MAXPGPATH];
426438

427439
#ifndefWIN32
440+
pgpid_tpm_pid;
441+
442+
/* Flush stdio channels just before fork, to avoid double-output problems */
443+
fflush(stdout);
444+
fflush(stderr);
445+
446+
pm_pid=fork();
447+
if (pm_pid<0)
448+
{
449+
/* fork failed */
450+
write_stderr(_("%s: could not start server: %s\n"),
451+
progname,strerror(errno));
452+
exit(1);
453+
}
454+
if (pm_pid>0)
455+
{
456+
/* fork succeeded, in parent */
457+
returnpm_pid;
458+
}
459+
460+
/* fork succeeded, in child */
428461

429462
/*
430463
* Since there might be quotes to handle here, it is easier simply to pass
431-
* everything to a shell to process them.
432-
*
433-
* XXX it would be better to fork and exec so that we would know the child
434-
* postmaster's PID directly; then test_postmaster_connection could use
435-
* the PID without having to rely on reading it back from the pidfile.
464+
* everything to a shell to process them. Use exec so that the postmaster
465+
* has the same PID as the current child process.
436466
*/
437467
if (log_file!=NULL)
438-
snprintf(cmd,MAXPGPATH,"\"%s\" %s%s < \"%s\" >> \"%s\" 2>&1 &",
468+
snprintf(cmd,MAXPGPATH,"exec\"%s\" %s%s < \"%s\" >> \"%s\" 2>&1",
439469
exec_path,pgdata_opt,post_opts,
440470
DEVNULL,log_file);
441471
else
442-
snprintf(cmd,MAXPGPATH,"\"%s\" %s%s < \"%s\" 2>&1 &",
472+
snprintf(cmd,MAXPGPATH,"exec\"%s\" %s%s < \"%s\" 2>&1",
443473
exec_path,pgdata_opt,post_opts,DEVNULL);
444474

445-
returnsystem(cmd);
475+
(void)execl("/bin/sh","/bin/sh","-c",cmd, (char*)NULL);
476+
477+
/* exec failed */
478+
write_stderr(_("%s: could not start server: %s\n"),
479+
progname,strerror(errno));
480+
exit(1);
481+
482+
return0;/* keep dumb compilers quiet */
483+
446484
#else/* WIN32 */
447485

448486
/*
449-
*On win32 we don't use system(). So we don't need to use& (which would
450-
*be START /B on win32). However, we still call the shell (CMD.EXE) with
451-
*ittohandle redirection etc.
487+
*As with the Unix case, it's easiest to usethe shell (CMD.EXE) to
488+
*handle redirection etc. UnfortunatelyCMD.EXE lacks any equivalent of
489+
*"exec", so we don't gettofind out the postmaster's PID immediately.
452490
*/
453491
PROCESS_INFORMATIONpi;
454492

@@ -460,10 +498,15 @@ start_postmaster(void)
460498
exec_path,pgdata_opt,post_opts,DEVNULL);
461499

462500
if (!CreateRestrictedProcess(cmd,&pi, false))
463-
returnGetLastError();
464-
CloseHandle(pi.hProcess);
501+
{
502+
write_stderr(_("%s: could not start server: error code %lu\n"),
503+
progname, (unsigned long)GetLastError());
504+
exit(1);
505+
}
506+
/* Don't close command process handle here; caller must do so */
507+
postmasterProcess=pi.hProcess;
465508
CloseHandle(pi.hThread);
466-
return0;
509+
returnpi.dwProcessId;/* Shell's PID, not postmaster's! */
467510
#endif/* WIN32 */
468511
}
469512

@@ -472,15 +515,21 @@ start_postmaster(void)
472515
/*
473516
* Find the pgport and try a connection
474517
*
518+
* On Unix, pm_pid is the PID of the just-launched postmaster. On Windows,
519+
* it may be the PID of an ancestor shell process, so we can't check the
520+
* contents of postmaster.pid quite as carefully.
521+
*
522+
* On Windows, the static variable postmasterProcess is an implicit argument
523+
* to this routine; it contains a handle to the postmaster process or an
524+
* ancestor shell process thereof.
525+
*
475526
* Note that the checkpoint parameter enables a Windows service control
476527
* manager checkpoint, it's got nothing to do with database checkpoints!!
477528
*/
478529
staticPGPing
479-
test_postmaster_connection(booldo_checkpoint)
530+
test_postmaster_connection(pgpid_tpm_pid,booldo_checkpoint)
480531
{
481532
PGPingret=PQPING_NO_RESPONSE;
482-
boolfound_stale_pidfile= false;
483-
pgpid_tpm_pid=0;
484533
charconnstr[MAXPGPATH*2+256];
485534
inti;
486535

@@ -535,29 +584,27 @@ test_postmaster_connection(bool do_checkpoint)
535584
optlines[5]!=NULL)
536585
{
537586
/* File is complete enough for us, parse it */
538-
longpmpid;
587+
pgpid_tpmpid;
539588
time_tpmstart;
540589

541590
/*
542-
* Make sanity checks. If it's for a standalone backend
543-
* (negative PID), or the recorded start time is before
544-
* pg_ctl started, then either we are looking at the wrong
545-
* data directory, or this is a pre-existing pidfile that
546-
* hasn't (yet?) been overwritten by our child postmaster.
547-
* Allow 2 seconds slop for possible cross-process clock
548-
* skew.
591+
* Make sanity checks. If it's for the wrong PID, or the
592+
* recorded start time is before pg_ctl started, then
593+
* either we are looking at the wrong data directory, or
594+
* this is a pre-existing pidfile that hasn't (yet?) been
595+
* overwritten by our child postmaster. Allow 2 seconds
596+
* slop for possible cross-process clock skew.
549597
*/
550598
pmpid=atol(optlines[LOCK_FILE_LINE_PID-1]);
551599
pmstart=atol(optlines[LOCK_FILE_LINE_START_TIME-1]);
552-
if (pmpid <=0||pmstart<start_time-2)
553-
{
554-
/*
555-
* Set flag to report stale pidfile if it doesn't get
556-
* overwritten before we give up waiting.
557-
*/
558-
found_stale_pidfile= true;
559-
}
560-
else
600+
if (pmstart >=start_time-2&&
601+
#ifndefWIN32
602+
pmpid==pm_pid
603+
#else
604+
/* Windows can only reject standalone-backend PIDs */
605+
pmpid>0
606+
#endif
607+
)
561608
{
562609
/*
563610
* OK, seems to be a valid pidfile from our child.
@@ -567,9 +614,6 @@ test_postmaster_connection(bool do_checkpoint)
567614
char*hostaddr;
568615
charhost_str[MAXPGPATH];
569616

570-
found_stale_pidfile= false;
571-
pm_pid= (pgpid_t)pmpid;
572-
573617
/*
574618
* Extract port number and host string to use. Prefer
575619
* using Unix socket if available.
@@ -635,42 +679,23 @@ test_postmaster_connection(bool do_checkpoint)
635679
}
636680

637681
/*
638-
* The postmaster should create postmaster.pid very soon after being
639-
* started. If it's not there after we've waited 5 or more seconds,
640-
* assume startup failed and give up waiting. (Note this covers both
641-
* cases where the pidfile was never created, and where it was created
642-
* and then removed during postmaster exit.) Also, if there *is* a
643-
* file there but it appears stale, issue a suitable warning and give
644-
* up waiting.
682+
* Check whether the child postmaster process is still alive. This
683+
* lets us exit early if the postmaster fails during startup.
684+
*
685+
* On Windows, we may be checking the postmaster's parent shell, but
686+
* that's fine for this purpose.
645687
*/
646-
if (i >=5)
688+
#ifndefWIN32
647689
{
648-
structstatstatbuf;
690+
intexitstatus;
649691

650-
if (stat(pid_file,&statbuf)!=0)
651-
{
652-
if (errno!=ENOENT)
653-
write_stderr(_("\n%s: could not stat file \"%s\": %s\n"),
654-
progname,pid_file,strerror(errno));
655-
returnPQPING_NO_RESPONSE;
656-
}
657-
658-
if (found_stale_pidfile)
659-
{
660-
write_stderr(_("\n%s: this data directory appears to be running a pre-existing postmaster\n"),
661-
progname);
692+
if (waitpid((pid_t)pm_pid,&exitstatus,WNOHANG)== (pid_t)pm_pid)
662693
returnPQPING_NO_RESPONSE;
663-
}
664694
}
665-
666-
/*
667-
* If we've been able to identify the child postmaster's PID, check
668-
* the process is still alive. This covers cases where the postmaster
669-
* successfully created the pidfile but then crashed without removing
670-
* it.
671-
*/
672-
if (pm_pid>0&& !postmaster_is_alive((pid_t)pm_pid))
695+
#else
696+
if (WaitForSingleObject(postmasterProcess,0)==WAIT_OBJECT_0)
673697
returnPQPING_NO_RESPONSE;
698+
#endif
674699

675700
/* No response, or startup still in process; wait */
676701
#if defined(WIN32)
@@ -836,7 +861,7 @@ static void
836861
do_start(void)
837862
{
838863
pgpid_told_pid=0;
839-
intexitcode;
864+
pgpid_tpm_pid;
840865

841866
if (ctl_command!=RESTART_COMMAND)
842867
{
@@ -876,19 +901,13 @@ do_start(void)
876901
}
877902
#endif
878903

879-
exitcode=start_postmaster();
880-
if (exitcode!=0)
881-
{
882-
write_stderr(_("%s: could not start server: exit code was %d\n"),
883-
progname,exitcode);
884-
exit(1);
885-
}
904+
pm_pid=start_postmaster();
886905

887906
if (do_wait)
888907
{
889908
print_msg(_("waiting for server to start..."));
890909

891-
switch (test_postmaster_connection(false))
910+
switch (test_postmaster_connection(pm_pid,false))
892911
{
893912
casePQPING_OK:
894913
print_msg(_(" done\n"));
@@ -914,6 +933,12 @@ do_start(void)
914933
}
915934
else
916935
print_msg(_("server starting\n"));
936+
937+
#ifdefWIN32
938+
/* Now we don't need the handle to the shell process anymore */
939+
CloseHandle(postmasterProcess);
940+
postmasterProcess=INVALID_HANDLE_VALUE;
941+
#endif
917942
}
918943

919944

@@ -1585,7 +1610,7 @@ pgwin32_ServiceMain(DWORD argc, LPTSTR *argv)
15851610
if (do_wait)
15861611
{
15871612
write_eventlog(EVENTLOG_INFORMATION_TYPE,_("Waiting for server startup...\n"));
1588-
if (test_postmaster_connection(true)!=PQPING_OK)
1613+
if (test_postmaster_connection(postmasterPID,true)!=PQPING_OK)
15891614
{
15901615
write_eventlog(EVENTLOG_ERROR_TYPE,_("Timed out waiting for server startup\n"));
15911616
pgwin32_SetServiceStatus(SERVICE_STOPPED);
@@ -1606,10 +1631,9 @@ pgwin32_ServiceMain(DWORD argc, LPTSTR *argv)
16061631
{
16071632
/*
16081633
* status.dwCheckPoint can be incremented by
1609-
* test_postmaster_connection(true), so it might not start
1610-
* from 0.
1634+
* test_postmaster_connection(), so it might not start from 0.
16111635
*/
1612-
intmaxShutdownCheckPoint=status.dwCheckPoint+12;;
1636+
intmaxShutdownCheckPoint=status.dwCheckPoint+12;
16131637

16141638
kill(postmasterPID,SIGINT);
16151639

‎src/bin/pg_ctl/t/001_start_stop.pl

Lines changed: 7 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -26,8 +26,13 @@
2626
close CONF;
2727
command_ok(['pg_ctl','start','-D',"$tempdir/data",'-w' ],
2828
'pg_ctl start -w');
29-
command_ok(['pg_ctl','start','-D',"$tempdir/data",'-w' ],
30-
'second pg_ctl start succeeds');
29+
# sleep here is because Windows builds can't check postmaster.pid exactly,
30+
# so they may mistake a pre-existing postmaster.pid for one created by the
31+
# postmaster they start. Waiting more than the 2 seconds slop time allowed
32+
# by test_postmaster_connection prevents that mistake.
33+
sleep 3if ($windows_os);
34+
command_fails(['pg_ctl','start','-D',"$tempdir/data",'-w' ],
35+
'second pg_ctl start -w fails');
3136
command_ok(['pg_ctl','stop','-D',"$tempdir/data",'-w','-m','fast' ],
3237
'pg_ctl stop -w');
3338
command_fails(['pg_ctl','stop','-D',"$tempdir/data",'-w','-m','fast' ],

0 commit comments

Comments
 (0)
[8]ページ先頭
©2009-2025 Movatter.jp