Hi,
I happened to be updating our machine running our buildfarm animals, and
I noticed something quite strange - the machine was unexpectedly running
out of disk space, which is rather suspicious as it's running just the
regression tests :-/
After a bit of investigation, I found this:
# pwd /mnt/fulmar/build/HEAD/pgsql.build/src/pl/plpgsql/src/results
# ls -l
total 57089152
-rw-r--r--. 1 pgbuild pgbuild 714 Jan 1 00:27 plpgsql_call.out
-rw-r--r--. 1 pgbuild pgbuild 58458515597 Mar 17 13:54
plpgsql_control.out
Right - the plpgsql_control.out output has about 58GB, which is somewhat
excessive I guess. The reason is fairly simple:
NOTICE: 1..3: i = 1
NOTICE: 1..3: i = 2
NOTICE: 1..3: i = 3
NOTICE: 1..10 by 3: i = 1
NOTICE: 1..10 by 3: i = 4
NOTICE: 1..10 by 3: i = 7
NOTICE: 1..10 by 3: i = 10
NOTICE: 1..11 by 3: i = 1
NOTICE: 1..11 by 3: i = 4
NOTICE: 1..11 by 3: i = 7
NOTICE: 1..11 by 3: i = 10
NOTICE: reverse 10..0 by 3: i = 10
NOTICE: reverse 10..0 by 3: i = 7
NOTICE: reverse 10..0 by 3: i = 4
NOTICE: reverse 10..0 by 3: i = 1
NOTICE: 2147483620..2147483647 by 10: i = 2147483620
NOTICE: 2147483620..2147483647 by 10: i = 2147483630
NOTICE: 2147483620..2147483647 by 10: i = 2147483640
NOTICE: 2147483620..2147483647 by 10: i = -2147483646
NOTICE: 2147483620..2147483647 by 10: i = -2147483636
NOTICE: 2147483620..2147483647 by 10: i = -2147483626
... many more NOTICE messages ...
Looking at the plpgsql_call.out timestamp, this seems to be running
since January 1, which also correlates with the last results reported to
pgbuildfarm.
Does that remind some known issue to anyone? It seems a bit like an
undetected overflow, but I don't see why should the behavior change
suddenly (AFAICS there were no updates to the machine just before
January 1st).
I wonder if this might be related ICC 14.0.3, but again - it has been
like that for a very long time.
Ideas?
--
Tomas Vondra http://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services