test sql/src/test/BugTracker/Tests/mdb_starts_with_sql_debug_64.SF-1999354.py fails with "too many digits"

[monetdb-team]

Date: 2010-07-07 14:52:40 +0200
From: @drstmane
To: SQL devs <>
Version: 11.1.1 (Mar2011) [obsolete]
CC: @njnes

Last updated: 2011-04-28 16:25:33 +0200

Comment 14248

Date: 2010-07-07 14:52:40 +0200
From: @drstmane

User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-GB; rv:1.9.1.10) Gecko/20100622 Fedora/3.5.10-1.fc12 Firefox/3.5.10
Build Identifier:

test sql/src/test/BugTracker/Tests/mdb_starts_with_sql_debug_64.SF-1999354.py fails with
"
QUERY = SELECT f2(1,2,3);
ERROR = !ERROR: too many digits (3 > 2)
"

on all platform except Debian and Windows:

http://monetdb.cwi.nl/testing/projects/monetdb/Candidate/sql/.mTests103/GNU.64.64.d.1-Fedora12/src_test_BugTracker/mdb_starts_with_sql_debug_64.SF-1999354.err.00.html

Reproducible: Always

Steps to Reproduce:

see nightly testing results
or
run test yourself

Actual Results:

test fails, i.e., output differs from approved stable output

Expected Results:

test should work fine, i.e., output should match approved stable output

I'm not sure whether the fact that Debian and Windows are the only platforms without RDF support installed is related or pure coincidence...

Comment 14262

Date: 2010-07-07 23:25:26 +0200
From: @drstmane

The other difference between our Debian & Windows machines on the one side and all other machines on the other side is that Debian & Windows are single core (Debian) or run MonetDB on purpose single-threaded (--gdk_nr_threads=1 on Windows).

Comment 14379

Date: 2010-07-13 14:21:07 +0200
From: @drstmane

The test works fine when running single threaded (Mtest.py -n1 ...); hence, the bug might well be related to mitosis/mergetable or dataflow.

Comment 14464

Date: 2010-07-18 09:40:00 +0200
From: @drstmane

Updated URL is:

http://monetdb.cwi.nl/testing/projects/monetdb/Stable/sql/.mTests103/GNU.64.64.d.1-Fedora12/src_test_BugTracker/mdb_starts_with_sql_debug_64.SF-1999354.err.00.html

Comment 14550

Date: 2010-07-28 10:51:04 +0200
From: @drstmane

Since today, i.e., most probably since this fix of bug #2626
http://dev.monetdb.org/hg/MonetDB/rev/2b7eec81b411 ,
the error message does no longer occur:
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/sql/.mTests103/GNU.64.64.d.1-Fedora10/src_test_BugTracker/mdb_starts_with_sql_debug_64.SF-1999354.err.00.html

However, the output still differs (slightly) between multi-threaded & single-threaded execution:
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/sql/.mTests103/GNU.64.64.d.1-Fedora10/src_test_BugTracker/mdb_starts_with_sql_debug_64.SF-1999354.out.00.html

Comment 14835

Date: 2010-08-30 09:23:18 +0200
From: @sjoerdmullender

The Jun2010-SP2 version has been released.

Comment 15071

Date: 2010-10-19 15:49:55 +0200
From: @drstmane

With Jun2010, the multi-threaded output still differs from the single-threaded output; cf.,
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/sql/.mTests103/GNU.64.64.d.1-Fedora10/src_test_BugTracker/mdb_starts_with_sql_debug_64.SF-1999354.out.00.html

With Oct2010, output appears to be as expected on all non-Windows systems (both single- & multi-threaded).
On Windows, the test times out as reported in
http://bugs.monetdb.org/show_bug.cgi?id=2699

Comment 15674

Date: 2011-03-28 17:36:27 +0200
From: @sjoerdmullender

The Mar2011 version has been released.

Comment 15704

Date: 2011-03-30 13:02:38 +0200
From: @njnes

has been fixed (combination of fixes in the last weeks).

Comment 15745

Date: 2011-04-28 16:25:33 +0200
From: @sjoerdmullender

The Apr2011 release has been uploaded.