pgpool: Test: stabilize 006.memqcache regression test.

Test: stabilize 006.memqcache regression test.

This commit attempts to stabilize the test, which occasionally fails.
Inspecting the buildfarm log showed that test2 failed.

test failed in test case 2

In the test2 pgproto is used to drive the test. The test data
includes erroneous query "SELECT setseed(10)" to see if pgpool correctly
handles an error response. In the PostgreSQL log, the error response
definitely returned to pgpool.

19025 2026-10-04 10:36:06.988 UTC ERROR: setseed parameter 10 is out of allowed range [-1,1]

However, pgppol log said that shutdown sequence started before the error.

2026-10-04 10:36:06.450: main pid 18965: LOG: exit handler called (signal: 2)

I suspect pgproto was terminated by timeout command before processing
the error response, because 10:36:06.450 is 1 second past from pgproto
started (10:36:05.488) and pgproto was started by timeout command to
give the runtime limit up to 1 second.
So, let's make the timeout from 1 to 5 second.

Backpatch-through: v4.3

Branch
------
V4_6_STABLE

Details
-------
https://git.postgresql.org/gitweb?p=pgpool2.git;a=commitdiff;h=d73298e0f9e073780b2ef624dbec43f4543af9f3

Modified Files
--------------
src/test/regression/tests/006.memqcache/test.sh | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)