Обсуждение: cygwin64 assertion failure
lorikeet seems to be stuck running the parallel tests, after having failed an assertion. Here's an excerpt from the log: 2016-10-09 11:52:35.751 EDT [57fa67c3.1148:11] LOG: statement: alter table tenk1 set (parallel_workers = 4); 2016-10-09 11:52:35.753 EDT [57fa67c3.1148:12] LOG: statement: explain (verbose, costs off) select parallel_restricted(unique1) from tenk1 where stringu1 = 'GRAAAA' order by 1; 2016-10-09 11:52:35.756 EDT [57fa67c3.1148:13] LOG: statement: explain (costs off) select length(stringu1) from tenk1 group by length(stringu1); 2016-10-09 11:52:35.758 EDT [57fa67c3.1148:14] LOG: statement: select length(stringu1) from tenk1 group by length(stringu1); TRAP: FailedAssertion("!(vmq->mq_sender == ((void *)0))", File: "/home/andrew/bf64/root/HEAD/pgsql.build/../pgsql/src/backend/storage/ipc/shm_mq.c", Line: 221) 2016-10-09 11:53:54.260 EDT [57fa6795.2a7c:2] WARNING: worker took too long to start; canceled Since then almost the only thing on the log is repeated instances of that last message. cheers andrew
On Sun, Oct 9, 2016 at 10:23 PM, Andrew Dunstan <andrew@dunslane.net> wrote: > lorikeet seems to be stuck running the parallel tests, after having failed > an assertion. > It looks like this problem has been discussed before [1]. Shall we do what Tom has suggested there? [1] - https://www.postgresql.org/message-id/15344.1473974558%40sss.pgh.pa.us -- With Regards, Amit Kapila. EnterpriseDB: http://www.enterprisedb.com
Amit Kapila <amit.kapila16@gmail.com> writes: > On Sun, Oct 9, 2016 at 10:23 PM, Andrew Dunstan <andrew@dunslane.net> wrote: >> lorikeet seems to be stuck running the parallel tests, after having failed >> an assertion. > It looks like this problem has been discussed before [1]. Shall we do > what Tom has suggested there? > [1] - https://www.postgresql.org/message-id/15344.1473974558%40sss.pgh.pa.us Well, we now have *two* issues. One is why is the Assert failing. The other is why is the system failing to clean up afterwards --- you'd expect that an Assert in a parallel worker would not be a case that the developers never saw ;-). I'd suggest fixing the latter first. regards, tom lane