January 2008 - developers-list

[Monetdb-developers] Run Mtest on windows...
by Romulo Goncalves 08 Jan '08

08 Jan '08

I am running Mtest on windows 64 to fix a test that I submitted yesterday and I am getting the following error when I start my Mtest: E:\MonetDB\Current\sql\NT>RunMtest.bat -rq Mtest.py: ERROR: ' mserver5 "--config=E:\MonetDB\Current\MonetDB5\NT\etc\monetdb5.conf" --debug=10 --set "monet_mod_pat h=E:\MonetDB\Current\\sql\NT\src\backends\monet5\.libs;E:\MonetDB\Current\\sql\NT\src\backends\monet5;E:\MonetDB\Current\ MonetDB5\NT\lib\MonetDB5" --set "gdk_dbfarm=E:\MonetDB\Current\\sql\NT\dbfarm" --set "sql_logdir=E:\MonetDB\Current\\sql\ NT\sql_logs" --set "xquery_logdir=E:\MonetDB\Current\\sql\NT\xquery_logs" --set mapi_open=true --set xrpc_open=true --set mapi_port=38998 --set xrpc_port=43861 --set monet_prompt= --trace --dbname=mTests` failed: !WARNING: GDKlockHome: ignoring empty or invalid .gdk_lock. !WARNING: BBPdir: initializing BBP. # MonetDB server v5.3.0, based on kernel v1.21.0 # Serving database 'mTests' # Compiled for i686-pc-win32/64bit with 32bit OIDs dynamically linked # Copyright (c) 1993-2007 CWI, all rights reserved # Visit http://monetdb.cwi.nl/ for further information #warning: please don't forget to set your vault key! #(see E:\MonetDB\Current\MonetDB5\NT\etc\monetdb5.conf) !FATAL: MALException:addUser:user 'monetdb' already exists ! Exit 1 Exception in thread Thread-12 (most likely raised during interpreter shutdown): Traceback (most recent call last): File "C:\Python25\lib\threading.py", line 460, in __bootstrap File "C:\Python25\lib\threading.py", line 623, in run File "C:\Python25\lib\threading.py", line 364, in wait File "C:\Python25\lib\threading.py", line 238, in wait <type 'exceptions.ValueError'>: list.remove(x): x not in list E:\MonetDB\Current\sql\NT> I already cleaned my dbfarm was you can see... Any idea? Regards, Romulo

2 2

[Monetdb-developers] find temp tables....
by Romulo Goncalves 05 Jan '08

05 Jan '08

In sql we have some problems in locate temp tables. If I create the following table: create global temporary t1(id int); If I do: select * from t1; or select * from tmp.t1; I can see the table. However, When I do alter table, copy into, create trigger, create index statements I cannot detect the table even if I use tmp.t1 Should I open a bug? Or there is a special reason for this? Regards, Romulo

2 1

[Monetdb-developers] commit options for create table statement....
by Romulo Goncalves 05 Jan '08

05 Jan '08

Reading the MonetDB code seems that commit option ON COMMIT DROP ON COMMIT PRESERVE ROWS ON COMMIT DELETE ROWS should not be used for normal tables (the ones which are not temporary). If that is the rule, the commit option should be allowed in the create table statement. If it is not the rule, the commit option is not being used at all for the normal tables. Look to the following example (auto-commit mode is on): >CREATE TABLE test1(id int) ON COMMIT DELETE ROWS; &3 >insert into test1 values(1); &2 1 -1 >select * from test1; &1 3 1 1 1 % sys.test1 # table_name % id # name % int # type % 1 # length [ 1 ] > So what is the rule for MonetDB? Regards, Romulo

2 1

[Monetdb-developers] Geom II
by Stefan de Konink 04 Jan '08

04 Jan '08

And I probably should take the other part of the documentation out there more seriously. I guess this is 'suppose' to happen if you do *not* include geom in your (jfu) database. gdb --args /opt/monetdb/bin/mserver5 --dbinit="include sql;" --dbname=osm --dbfarm=/mnt/media/MonetDB5/dbfarm GNU gdb 6.7.1 Copyright (C) 2007 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "x86_64-pc-linux-gnu"... Using host libthread_db library "/lib/libthread_db.so.1". (gdb) run Starting program: /opt/monetdb/bin/mserver5 --dbinit=include\ sql\; --dbname=osm --dbfarm=/mnt/media/MonetDB5/dbfarm warning: no loadable sections found in added symbol-file system-supplied DSO at 0x7fff30ffd000 [Thread debugging using libthread_db enabled] [New Thread 0x2b887c85d860 (LWP 2919)] [New Thread 0x40800950 (LWP 2922)] # MonetDB server v5.3.0, based on kernel v1.21.0 # Serving database 'osm' # Compiled for x86_64-unknown-linux-gnu/64bit with 64bit OIDs dynamically linked # Copyright (c) 1993-2007 CWI, all rights reserved # Visit http://monetdb.cwi.nl/ for further information #warning: please don't forget to set your vault key! #(see /opt/monetdb/etc/monetdb5.conf) [New Thread 0x41001950 (LWP 2923)] # Listening for connection requests on mapi:monetdb://127.0.0.1:50000/ # MonetDB/SQL module v2.21.0 loaded >!ERROR: BATnew:tt error Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 0x41001950 (LWP 2923)] 0x00002aaab26c8d3c in temp_create (b=0x0) at bat_utils.mx:130 130 temp_dup(b->batCacheid); (gdb) bt #0 0x00002aaab26c8d3c in temp_create (b=0x0) at bat_utils.mx:130 #1 0x00002aaab26c4d76 in e_bat (type=-31) at bat_storage.mx:84 #2 0x00002aaab26c62ee in load_bat (sname=0xba3028 "sys", tname=0x10607d8 "nodes_geom", bname=0x107d118 "geometry", type=-31, sz=1024, cnt=0x1060700) at bat_storage.mx:454 #3 0x00002aaab26c6660 in create_col (tr=0xba2ee8, c=0x107d088) at bat_storage.mx:518 #4 0x00002aaab26b639b in load_column (tr=0xba2ee8, t=0x10606c8, rid=193) at store.mx:516 #5 0x00002aaab26b691c in load_table (tr=0xba2ee8, s=0xba2f58, rid=193) at store.mx:583 #6 0x00002aaab26b7dd3 in load_schema (tr=0xba2ee8, id=1128, rid=38) at store.mx:850 #7 0x00002aaab26b819e in load_trans (tr=0xba2ee8, id=1128) at store.mx:902 #8 0x00002aaab26ba16d in store_init (debug=0, store=store_bat, logdir=0x63cfc0 "/opt/monetdb/var/MonetDB5/sql_logs", dbname=0x63ce90 "osm", stk=0) at store.mx:1328 #9 0x00002aaab269707d in mvc_init (debug=0, store=store_bat, stk=0) at sql_mvc.mx:247 #10 0x00002aaab26367bb in SQLinit (c=0x6046c8) at sql_scenario.mx:227 #11 0x00002b8879e003af in initScenario (c=0x6046c8, s=0x2b887a01b070) at mal_scenario.mx:263 #12 0x00002b8879e0159c in setScenario (c=0x6046c8, nme=0xad8123 "sql") at mal_scenario.mx:471 #13 0x00002b8879dc9951 in MSscheduleClient (command=0xad80e8 "LIT", challenge=0x41000fc0 "Roh8IvYDJa", fin=0xacdf40, fout=0xacdea0) at mal_session.mx:342 #14 0x00002aaab0f8527b in doChallenge (cmd=0x0, in=0xacdd60, out=0xacdcc0) at mserver.mx:481 #15 0x00002aaab0f85533 in SERVERlistenThread (Sock=0x9f0938) at mserver.mx:563 #16 0x00002b887b8d5067 in start_thread () from /lib/libpthread.so.0 #17 0x00002b887c5d9bed in clone () from /lib/libc.so.6 People may wonder why I tried anyway... The simple query "select * from nodes LIMIT 1;" also took far more time (22s) to complete than expected. Stefan

1 0

[Monetdb-developers] Geom and its examples.
by Stefan de Konink 04 Jan '08

04 Jan '08

Hi, So I decided to try out Geom, and after copying in a small set of data. I decided to test if it was any faster than my own made up SQL fantasy, which greatly seems to depend on the state MonetDB is in, I had some nice results before my Geom test, but I guess I just have overloaded the database again. select * from nodes, node_tags where latitude >= 52.375 AND latitude <= 52.379 AND longitude >= 4.89 AND longitude <= 4.90 AND nodes.id = node_tags.node AND name='amenity' and value='atm'; Took before 8 seconds... now it is ending nowhere. (In the over 10 minutes range.) Next to this now is 27.8GB in virtual memory (my entire dbfarm directory, while the data that is required for this query BY FAR not that amount of data.) So I suspect a problem in the Geom code :) read on :) I'm using the very latest CVS HEAD. sql>SELECT id, geometry FROM nodes_geom LIMIT 1; +----+---------------------------------------------------------------------+ | id |geometry | +====+=====================================================================+ | 2 |POINT (50.1360073999999969 8.3023717000000001) | +----+---------------------------------------------------------------------+ Looks shining. Lets do a basic operation: sql>SELECT id, X(geometry) FROM nodes_geom LIMIT 1; !types point(0,0) (wkb) and geometry(0,0) (wkb) are not equal Hum. Was it my poor choose of column name? So I decided to try out the example: http://monetdb.cwi.nl/projects/monetdb/SQL/Documentation/SQL_002fGIS.html sql>SELECT forests.name,buildings.name more>FROM forests,buildings more>WHERE forests.name = 'Green Forest' and more> Overlaps(forests.shape, buildings.outline) = true; !types multipolygon(0,0) (wkb) and geometry(0,0) (wkb) are not equal So what went so terribly wrong? Or better: what can be done to fix it. Yours Sincerely, Stefan de Konink

1 0

Re: [Monetdb-developers] [Monetdb-checkins] MonetDB/NT monetdb_winconfig.h, , 1.6, 1.7 winrules.msc, , 1.16, 1.17
by Romulo Goncalves 03 Jan '08

03 Jan '08

Stefan Manegold wrote: > Update of /cvsroot/monetdb/MonetDB/NT > In directory sc8-pr-cvs16.sourceforge.net:/tmp/cvs-serv15904/NT > > Modified Files: > monetdb_winconfig.h winrules.msc > Log Message: > > made OID size (command-line-) configurable (also) on 64-bit Windows: > As opposed to Unix, where the default OID size is identical to the > native machine word size (in particular 64-bit OIDs on 64-bit systems), > the default OID size on 64-bit Windows has been and remains 32-bit > (see NT/monetdb_winconfig.h for details). > This default can be overruled by compiling with `nmake MONET_OID64=1`. > (In case the default in NT/monetdb_winconfig.h is changed to 64-bit > OIDs on 64-bit Windows, 32-bit OIDs on 64-bit Windows can be enabled > by compiling with `nmake MONET_OID32=1`.) Following these rules I decided to compile MonetDB on windows like this: E:\MonetDB\Current\MonetDB\NT>nmake MONET_OID64=1 Microsoft (R) Program Maintenance Utility Version 8.00.50727.42 Copyright (C) Microsoft Corporation. All rights reserved. "E:\Microsoft Visual Studio 8\VC\BIN\amd64\nmake.exe" /nologo /f ".\..\Makefile.msc" "prefix=E:\MonetDB\Current\M onetDB\NT" "bits=32" all if not exist "src" mkdir "src" copy ".\..\src\Makefile.msc" "src\Makefile" 1 file(s) copied. cd "src" && "E:\Microsoft Visual Studio 8\VC\BIN\amd64\nmake.exe" /nologo "prefix=E:\MonetDB\Current\MonetDB\NT" "bits=32" all if not exist "common" mkdir "common" copy ".\..\..\src\common\Makefile.msc" "common\Makefile" 1 file(s) copied. cd "common" && "E:\Microsoft Visual Studio 8\VC\BIN\amd64\nmake.exe" /nologo "prefix=E:\MonetDB\Current\MonetDB\N T" "bits=32" all cl -GF -W3 -wd4273 -wd4102 -MD -nologo -Ox -I. -I.\..\.. "-IC:\Pthreads\include" -I.\..\..\..\src\common -DHAVE_L IBZ "-IC:\zlib-1.2.3.win32\include" -DHAVE_ICONV "-IC:\iconv-1.9.2.win32\include" -DLIBSTREAM -DCOMPILE_DL_STREAM - Fo"stream.obj" -c "stream.c" stream.c As you can see the compilation is done for 32 bits and I would like to do it for 64 bits. Could someone give a help here please. Thanks in advance, Romulo

2 1

Re: [Monetdb-developers] [Monetdb-pf-checkins] pathfinder/modules/pftijah pftijah.mx, , 1.157, 1.158
by Stefan Manegold 02 Jan '08

02 Jan '08

On Tue, Dec 11, 2007 at 10:08:13AM +0000, Jan Flokstra wrote: > Update of /cvsroot/monetdb/pathfinder/modules/pftijah > In directory sc8-pr-cvs16.sourceforge.net:/tmp/cvs-serv18823 > > Modified Files: > pftijah.mx > Log Message: > - add an optional extra commit after document indexing to survey a bug > where a server restart leads to a zero filled index bat. Just a stupid(?) question: In the code below, the "extra_commit" is initialized to "false", and never changed --- hence, the extra subcommit() in if ( extra_commit ) subcommit(_tj_collection_str(ftiName)); is never called, right? Hence, what is/was the intention/need of this checkin? Stefan > Index: pftijah.mx > =================================================================== > RCS file: /cvsroot/monetdb/pathfinder/modules/pftijah/pftijah.mx,v > retrieving revision 1.157 > retrieving revision 1.158 > diff -u -d -r1.157 -r1.158 > --- pftijah.mx 4 Oct 2007 10:41:12 -0000 1.157 > +++ pftijah.mx 11 Dec 2007 10:08:09 -0000 1.158 > @@ -315,7 +315,9 @@ > # # > ##################################################################### > > -var verbose := false; > +var verbose := false; > + > +var extra_commit := false; > > const dflt_ft_index := "DFLT_FT_INDEX"; > const dflt_bg_index := "DFLT_FT_INDEX"; > @@ -997,6 +999,7 @@ > _tj_add2collection(ftiName, collBat, uri_loc, uri_name, store); > _tj_finalize_collection(ftiName, collBat, FALSE); > _tj_commit(collBat); > + if ( extra_commit ) subcommit(_tj_collection_str(ftiName)); > if ( timing ) { > var ms := (usec()-t_start)/1000; > printf("#C[%s]:tj_add2collection(): + aggregate time = %lld.%03llds.\n",ftiName,/(ms,1000),%(ms,1000)); > @@ -1038,6 +1041,7 @@ > printf("#C[%s]:tj_add2collection(BAT): + aggregate time = %lld.%03llds.\n",ftiName,/(ms,1000),%(ms,1000)); > } > _tj_commit(collBat); > + if ( extra_commit ) subcommit(_tj_collection_str(ftiName)); > }); > lock_unset(coll_lock); > if (not(isnil(err))) ERROR(err); > > > ------------------------------------------------------------------------- > SF.Net email is sponsored by: > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Monetdb-pf-checkins mailing list > Monetdb-pf-checkins(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/monetdb-pf-checkins -- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 |

2 1