Re: Vacuum full hang => strace - Mailing list pgsql-admin
From | Gaetano Mendola |
---|---|
Subject | Re: Vacuum full hang => strace |
Date | |
Msg-id | b1ld9b$1lu9$1@news.hub.org Whole thread Raw |
In response to | Re: Vacuum full hang (<mallah@trade-india.com>) |
Responses |
Re: Vacuum full hang => strace
Re: Vacuum full hang => strace |
List | pgsql-admin |
<mallah@trade-india.com> wrote in message news:56131.203.122.61.30.1044193279.squirrel@mail.trade-india.com... > > are u sure VACUUM FULL is hanging? > > run in verbose mode... > > VACUUM FULL VERBOSE ANALYZE <tablename> I don't see nothing going on > and you are really desperate get the pid of the > backend who is doing the vacuum and > > strace -p <pid> i am sure u will see it spitting lots of read and writes. > ;-) I see only: recv(9, "Qvacuum full;\0", 8192, 0) = 14 send(5, "\3\0\0\0%\0\0\0!\0\0\0wy\0\0\33H\26\0d\0\0\0vacuum f"..., 37, 0) = 37 time(NULL) = 1044264831 time([1044264831]) = 1044264831 getpid() = 31095 rt_sigaction(SIGPIPE, {0x403248c8, [], 0x4000000}, {SIG_IGN}, 8) = 0 send(8, "<135>Feb 3 10:33:51 postgres[31"..., 79, 0) = 79 rt_sigaction(SIGPIPE, {SIG_IGN}, NULL, 8) = 0 gettimeofday({1044264831, 553516}, NULL) = 0 time(NULL) = 1044264831 time([1044264831]) = 1044264831 getpid() = 31095 rt_sigaction(SIGPIPE, {0x403248c8, [], 0x4000000}, {SIG_IGN}, 8) = 0 send(8, "<135>Feb 3 10:33:51 postgres[31"..., 75, 0) = 75 rt_sigaction(SIGPIPE, {SIG_IGN}, NULL, 8) = 0 time(NULL) = 1044264831 time([1044264831]) = 1044264831 getpid() = 31095 rt_sigaction(SIGPIPE, {0x403248c8, [], 0x4000000}, {SIG_IGN}, 8) = 0 send(8, "<135>Feb 3 10:33:51 postgres[31"..., 84, 0) = 84 rt_sigaction(SIGPIPE, {SIG_IGN}, NULL, 8) = 0 open("/var/lib/pgsql/data/global/pgstat.stat", O_RDONLY) = 37 fstat64(37, {st_mode=S_IFREG|0600, st_size=52815, ...}) = 0 old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x4001f000 read(37, "D\1\0\0\0\340\2662\10\0\0\0\0\216\6\0\0\0\0\0\0S\24\0\0"..., 4096) = 4096 brk(0x83c8000) = 0x83c8000 read(37, "\0\0\0\0\0\0:\0\0\0\0\0\0\0\335*\224\1\0\0\0\0\217\353"..., 4096) = 4096 brk(0x83cd000) = 0x83cd000 read(37, "\0\0\0\0\0\0\350\22\250\2\0\0\0\0$\243\247\2\0\0\0\0\0"..., 4096) = 4096 read(37, "\0\0\0\0\0\0\25\4\0\0\0\0\0\0\0\0\0\0TS\352*\0\0\0\0\0"..., 4096) = 4096 brk(0x83d6000) = 0x83d6000 read(37, "\0\0\0\0\0\0\0\0\0\0T\204\352*\0\0\0\0\0\0\0\0\0\0\0\0"..., 4096) = 4096 read(37, "\0\0T\256H\26\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 4096) = 4096 read(37, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 4096) = 4096 read(37, "\0\0\0\0\0\0\0\24\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 4096) = 4096 brk(0x83e7000) = 0x83e7000 read(37, "\0\0\0\0\0\0\0\24\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 4096) = 4096 read(37, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\1\0\0\0"..., 4096) = 4096 brk(0x83f0000) = 0x83f0000 read(37, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0B\33H\26\0}\0\0\0<"..., 4096) = 4096 read(37, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 4096) = 4096 read(37, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 4096) = 3663 close(37) = 0 munmap(0x4001f000, 4096) = 0 lseek(25, 0, SEEK_END) = 8192 lseek(3, 0, SEEK_END) = 376832 gettimeofday({1044264831, 559402}, NULL) = 0 setitimer(ITIMER_REAL, {it_interval={0, 0}, it_value={1, 0}}, {it_interval={0, 0}, it_value={0, 0}}) = 0 semop(21037068, 0xbfffed00, 1) = -1 EINTR (Interrupted system call) --- SIGALRM (Alarm clock) --- sigreturn() = ? (mask now []) semop(21037068, 0xbfffed00, 1 that's all, after this remain there for ever. Is true that in other and some people here use to connect to DB throw a JDBC interface and with the parameter Autocommit ON there is always a process inside a transaction.... Ciao Gaetano.
pgsql-admin by date: