Re: missing comment in lwlock.c
От | Andres Freund |
---|---|
Тема | Re: missing comment in lwlock.c |
Дата | |
Msg-id | 20160801153913.7tbgiyqsjqwmeivl@alap3.anarazel.de обсуждение исходный текст |
Ответ на | missing comment in lwlock.c (Jens.Wilke@parcIT.de) |
Ответы |
Re: missing comment in lwlock.c
|
Список | pgsql-bugs |
Hi, On 2016-08-01 10:43:43 +0200, Jens.Wilke@parcIT.de wrote: > there is no comment on "PANIC: queueing for lock while waiting on another > one" in lwlock.c That's because it's not expected to be ever reached. > Postgres Version is 9.5.3. > > 2016-07-25 18:32:54 CEST [45824]: [185-1] > user=svc_sqlbalancer_worker,db=adr,client=::1 FATAL: *semop(id=6389799) > failed: Identifier removed* > 2016-07-25 18:32:54 CEST [45824]: [186-1] > user=svc_sqlbalancer_worker,db=adr,client=::1 CONTEXT: SQL statement > "create temp table ausfall_ereignis_tmp as > 2016-07-25 18:32:54 CEST [45824]: [187-1] > user=svc_sqlbalancer_worker,db=adr,client=::1 STATEMENT: SELECT * FROM > svc_sqlbalancer.fx_run_work(50119) > 2016-07-25 18:32:54 CEST [45824]: [188-1] > user=svc_sqlbalancer_worker,db=adr,client=::1 *PANIC: queueing for lock > while waiting on another one* > > 2016-07-25 18:32:55 CEST [12024]: [176-1] user=,db=,client= LOG: > server process (PID 45824) was terminated by signal 6: Aborted > 2016-07-25 18:33:16 CEST [12024]: [179-1] user=,db=,client= LOG: all > server processes terminated; reinitializing > 2016-07-25 18:33:16 CEST [12024]: [180-1] user=,db=,client= LOG: could > not remove shared memory segment "/PostgreSQL.605661915": No such file or > directory > 2016-07-25 18:33:16 CEST [12024]: [181-1] user=,db=,client= LOG: > semctl(5144577, 0, IPC_RMID, ...) failed: Invalid argument > 2016-07-25 18:33:16 CEST [12024]: [182-1] user=,db=,client= LOG: > semctl(5177346, 0, IPC_RMID, ...) failed: Invalid argument > 2016-07-25 18:33:16 CEST [12024]: [183-1] user=,db=,client= LOG: > semctl(5210115, 0, IPC_RMID, ...) failed: Invalid argument This sounds like it's a consequence of sysv semaphores and shared memory being removed while postgres is running. Any chance this was a postgres running on a systemd machine with RemoveIPC=true? Andres
В списке pgsql-bugs по дате отправления: