On 01/11/11 02:46, Alexander wrote:
> Hi all.
>
> I've encountered the following problem: pg_restore hangs while
> restoring LOBs. The situation is as follows:
> 1) TARed dump, created by pg_dump 8.3.6
> 2) pg_restore from Postgresql 9.1.1
> 3) pg_restore restores schema and most of LOBs, but hangs on one of
> them (the same LOB every time). 'ps' shows 'R+' state for
> pg_restore, and GDB (attached to pg_restore) gives the following:
>
> (gdb) bt
> #0 0x0000003ec6ec6750 in __read_nocancel () from /lib64/libc.so.6
> #1 0x0000003ec6e6cd57 in _IO_new_file_underflow () from /lib64/libc.so.6
> #2 0x0000003ec6e6ba4e in _IO_file_xsgetn_internal () from /lib64/libc.so.6
> #3 0x0000003ec6e61dd2 in fread () from /lib64/libc.so.6
> #4 0x000000000040d0e2 in PQconnectdbParams ()
> #5 0x000000000040d7fd in PQconnectdbParams ()
> #6 0x000000000040e78a in PQconnectdbParams ()
> #7 0x0000000000408035 in PQconnectdbParams ()
> #8 0x0000000000409116 in PQconnectdbParams ()
> #9 0x0000000000403b91 in PQconnectdbParams ()
> #10 0x0000003ec6e1d994 in __libc_start_main () from /lib64/libc.so.6
> #11 0x0000000000402f99 in PQconnectdbParams ()
> #12 0x00007fff51ae6188 in ?? ()
> #13 0x0000000000000000 in ?? ()
>
> Does anyone know, what can be wrong?
Disk issues? Have you checked dmesg, the SMART status of your disks, and
syslog?
Can you copy the dump with the "cp" command?
Can you use "pg_restore" to produce a .sql dump from it?
--
Craig Ringer