October 2009 - users-list

[MonetDB-users] prepared exec did not translate date string
by Allen Zhang 16 Oct '09

16 Oct '09

Hi, there, We are testing the Aug SP1 (v5.14.2) release, it seems the exec did not translate the date string, please run attached sql against Aug SP1, it give out the following error, the same script works file on mserver v5.10.4 (May SP1). Sounds like a bug to me. 15:19:53 [CREATE - 0 row(s), 0.000 secs] Command processed. No rows were affected 15:19:53 [SELECT - 3 row(s), 0.000 secs] Result set fetched 15:19:54 [PREPARE - 1 row(s), 0.000 secs] Result set fetched 15:19:54 [EXEC - 0 row(s), 0.000 secs] SQLException:sql.prepare:wrong argument for prepared statement (expected date instead of char) 15:19:54 [EXEC - 0 row(s), 0.000 secs] current transaction is aborted (please ROLLBACK) 15:19:54 [EXEC - 0 row(s), 0.000 secs] current transaction is aborted (please ROLLBACK) 15:19:54 [EXEC - 0 row(s), 0.000 secs] current transaction is aborted (please ROLLBACK) 15:19:54 [COMMIT - 0 row(s), 0.000 secs] COMMIT: failed ERROR: COMMIT: transaction is aborted, will ROLLBACK instead 15:19:54 [SELECT - 3 row(s), 0.015 secs] Result set fetched 15:19:54 [DROP - 0 row(s), 0.000 secs] Command processed. No rows were affected ... 10 statement(s) executed, 7 row(s) affected, exec/fetch time: 0.015/0.000 sec [3 successful, 2 warnings, 5 errors] Thanks for the attention. Allen

2 1

[MonetDB-users] BATpropcheck exception
by Alex Bo. 15 Oct '09

15 Oct '09

Hi, I haven't been able to reproduce this, but when running an INSERT query I got: !MALEXCEPTION: !ERROR: BATpropcheck: ... tuples seqbase of dense oid bat is wrong! 0 != 12306 I am running Aug09 SP1 compiled from CVS on Windows. What could have caused this error? Thanks. -- View this message in context: http://www.nabble.com/BATpropcheck-exception-tp25909476p25909476.html Sent from the monetdb-users mailing list archive at Nabble.com.

1 0

[MonetDB-users] Use cases for seperate servers vs single database
by Stefan de Konink 11 Oct '09

11 Oct '09

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hi, What is the developer perspective of running a single database with a single mapi interface but different users (and rights) versus multiple databases having each one user. Is there a good reason not to merge databases? I understand that each MonetDB instance thinks it can access all the memory on the system available, that would indicate it might be better to run a single server. Could someone elaborate on the above? Stefan -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.11 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEAREKAAYFAkrQsBgACgkQYH1+F2Rqwn3olACffKnjqdhef+fXkmKwmdOLvJkK ltgAnAw/MIIBHfXokLAbV3sA9RMUG/Hw =2fGw -----END PGP SIGNATURE-----

2 2

[MonetDB-users] MonetDB/XQuery with Java/Perl
by listanand＠gmail.com 09 Oct '09

09 Oct '09

Hi all, I am new to MonetDB/XQuery, and have a question to ask. Actually, I am new to MonetDB/XQuery, XML, SOAP, RPC, and XQuery also, and so you can very well imagine how overwhelmed I must be at this point :( I do not need super efficient query processing, but what I do certainly need is a convenient hassle free programmatic access to the XQuery database. I looked at the documentation and it appears that the Perl and Java APIs still need some work, and are not ready for prime time. Is that correct? Also, do I need to learn SOAP, RPC etc to be able to issue XQueries from my program? Thanks in advance. Sorry if I seem grumpy (I am!). Andy

1 0

[MonetDB-users] mapilib latin1
by wilfried＠hafner.ws 09 Oct '09

09 Oct '09

Hi! I have a table with a varchar(100) column, when I connect to the server with mclient.bat -Elatin1 option I'm able to insert and display characters like ä and ö correctly. But I have problems to do the same with a c++ app and the mapilib. Does anyone have some hints? regards, Wilfried

2 2

[MonetDB-users] Help with table sizes
by david85 08 Oct '09

08 Oct '09

Hello, I am currently trying to figure out how I can determine the size, i.e. the disk space, of the different tables. I know that they are stored in the dbfarm/database_name/bat folder, but I don't know which folder holds the data for which table. I looked at the manual but the bbp.getDiskSpace command didn't work (or maybe I used it wrong). I have also searched for this topic, but there was no concrete answer. Thank you in advance. Best Regards -- View this message in context: http://www.nabble.com/Help-with-table-sizes-tp25728470p25728470.html Sent from the monetdb-users mailing list archive at Nabble.com.

2 2

[MonetDB-users] Support for ANSI 1992/99 BIT
by McKennirey.Matthew 08 Oct '09

08 Oct '09

Although ANSI 2003 eliminated the BIT data type, I was wondering if anyone has developed support for this in MonetDB. Specifically we would like to store UUIDs in a column, as a 128 bit BIT type. Storing a UUID in its canonical string form (36 characters = 288 bits) seems inefficient. We would also like to consider using bit masks for access control. Thanks

1 0

[MonetDB-users] mapilib on windows
by wilfried＠hafner.ws 08 Oct '09

08 Oct '09

Hello! I have a question on getting the mapi working on windows... I downloaded the MonetDB-client-1.32.2.zip from the sources directory. I tried now to open the solution file in the NT directory but its not possible, I use vc++ 2008 express. What version of vc++ should I use? Is there an existing .lib in one of the other downloadable packages? I would like to create a sql driver for qt4 in c++... Kind Regards, Wilfried

2 2

[MonetDB-users] Insert slowdown
by Guillaume Theoret 07 Oct '09

07 Oct '09

Hi, I have a question about insert performance. At first, with an empty table, it took a little over 1 second (around 1.5) to insert 1000 rows. Now, with 5.2MM rows, it's taking ~65s. Is this to be expected? Thanks, Guillaume

2 4

[MonetDB-users] Segmentation fault on insert
by Guillaume Theoret 07 Oct '09

07 Oct '09

This has happened twice in two days so far. All I do is insert rows in a loop and eventually I hit a segmentation fault. [...] 160000 done - 2.1121% in 432.97 (0.0026 avg) 161000 done - 2.1253% in 435.68 (0.0027 avg) 162000 done - 2.1385% in 438.28 (0.0026 avg) 163000 done - 2.1517% in 442.37 (0.0041 avg) Segmentation fault monetdb@dev3:~$ MonetDB/bin/monetdb status name state uptime health last crash test running 2m 56s 34%, 0s 2009-10-06 15:45:01 Is there any way to recover from this? I tried stopping the db but it would just hang. I tried restarting monetdb/merovingian and starting the db again but it said this: starting database 'test'... FAILED: starting 'test' failed: database 'test' has inconsistent state (running but dead), review merovingian's logfile for any peculiarities This seems to be the log from when the crash occured: 2009-10-07 12:47:20 MSG merovingian[3081]: client localhost:40950 has disconnected from proxy 2009-10-07 12:47:20 MSG merovingian[3081]: client localhost:38976 has disconnected from proxy 2009-10-07 12:47:20 MSG merovingian[3081]: database 'test' (3865) was killed by signal 11 2009-10-07 14:03:29 MSG merovingian[3081]: database 'test' has crashed after start on 2009-10-06 15:45:01, attempting restart, up min/avg/max: 0s/0s/0s, crash average: 1.00 0.20 0.07 (2-0=2) 2009-10-07 14:03:30 MSG test[6042]: arguments: /home/guillaume/MonetDB/bin/mserver5 --config=/home/guillaume/MonetDB/etc/monetdb5.conf --dbnam e=test --dbinit=include sql; --set monet_daemon=yes --set mapi_open=false --set mapi_autosense=true --set mapi_port=50001 --set monet_vault_key=/home/guillaume/MonetDB/var/MonetDB5/dbfarm/test/.vaultkey And this is the log when I tried to start it: 2009-10-07 14:08:39 ERR merovingian[6059]: failed to fork mserver: database 'test' has inconsistent state (running but dead), review merovingian's logfile for any peculiarities If I recompile with debug mode enabled will it dump a stack trace somewhere when it segfaults? Thanks, Guillaume

1 1