Обсуждение: BUG #1549: initdb doesn't work
The following bug has been logged online: Bug reference: 1549 Logged by: Daniel van Eeden Email address: daniel_e@dds.nl PostgreSQL version: 8.0.1 Operating system: Debian GNU/Linux for SPARC (sarge) Description: initdb doesn't work Details: creating template1 database in /opt/postgresql/data/base/1 ... FATAL: semctl(10977333, 16, SETVAL, 536) failed: Invalid argument child process exited with exit code 1 initdb: removing data directory "/opt/postgresql/data" Postgresql is compiled from the official sources in 64-bit mode.
Daniel van Eeden wrote: > > The following bug has been logged online: > > Bug reference: 1549 > Logged by: Daniel van Eeden > Email address: daniel_e@dds.nl > PostgreSQL version: 8.0.1 > Operating system: Debian GNU/Linux for SPARC (sarge) > Description: initdb doesn't work > Details: > > creating template1 database in /opt/postgresql/data/base/1 ... FATAL: > semctl(10977333, 16, SETVAL, 536) failed: Invalid argument > child process exited with exit code 1 > initdb: removing data directory "/opt/postgresql/data" That is a problem with semaphores in your kernel. Have you modified your postgresql.conf from the default? If so, you should read the PostgreSQL documentation on how to modify your kernel for larger settings. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073
Bruce Momjian <pgman@candle.pha.pa.us> writes: > Daniel van Eeden wrote: >> creating template1 database in /opt/postgresql/data/base/1 ... FATAL: >> semctl(10977333, 16, SETVAL, 536) failed: Invalid argument > That is a problem with semaphores in your kernel. Have you modified > your postgresql.conf from the default? If so, you should read the > PostgreSQL documentation on how to modify your kernel for larger > settings. But it's failing at semctl() not semget(). So it seems he hasn't run out of semaphores. I wonder if SEMVMX is unusually small on his setup (ie, less than 536). My Linux semctl man page says that that case would return ERANGE not EINVAL, but ... regards, tom lane