January 2016 - users-list

Issue with Perl monetdb DBD plugin for DBI on a VM
by Puthick Hok 22 Jan '16

22 Jan '16

Hi, I'm using MonetDB-11.19.7 compiled from source on openSUSE 12.1. I'm also using the Perl client code from MonetDB source to connect to MonetDB from Perl via Perl DBI module. It works fine for a physical server but a VM server with a vmdk drive. It has an issue that the database handler must be disconnected and reconnected in order to execute more SQL statements; otherwise there will be an error saying 'Statement handle not marked as Active'. So I have written 2 testing Perl scripts doing SQL select for 100 iterations. One script is to demonstrate that it will fail while the other will work for all the iterations. The one that works is the script that connect and disconnect to MonetDB in every iteration before executing the SQL. In order to help you replicate the problem, there is an attached tar.gz file containing the 2 perl scripts - one for success and the other one for failure. Due to the size limitation, I cannot include the database dump file for the demonstration. Regards, Puthick

2 1

Query timeout
by Luciano Sasso 21 Jan '16

21 Jan '16

Any idea, how to set query timeout ? Other question, the limit of 65 connection is for merogivian process or mserver ? Att, -- Luciano Sasso Vieira Data Scientist & Solutions Architect luciano(a)gsgroup.com.br <http://www.gsgroup.com.br> | tel: 17 3353-0833 | cel: 17 99706-9335 www.gsgroup.com.br <http://www.gsgroup.com.br> --- Este email foi escaneado pelo Avast antivírus. https://www.avast.com/antivirus

1 0

How to subtract two columns from two different datasets?
by integrity 20 Jan '16

20 Jan '16

Dear experts, I have columns coming from two different datasets which also resides in two different tables in the same layout,now i need the difference of that columns. Example: sql>select (select mag from one_to_many o, targets16 t where o.new_targetid=t.id) - (select mag from one_to_many o,targets16 t where o.old_uniqueid=t.id) ; cardinality violation (22>1) p.s. these two tables have both 22 rows. But if I subtract on single values, it works: select (select avg(mag) from mag1) - (select avg(mag) from mag2); +--------------------------+ | sql_sub_L1 | +==========================+ | -0.015908087452235264 | +--------------------------+ 1 tuple (4.486ms) can any one help me out? Thanks In Advance !!!! Meng

2 1

Multi-core join performance
by Alexandros Koliousis 20 Jan '16

20 Jan '16

Hi, I am joining two tables, each with 32K tuples. The query runs for ~23-24 seconds on my (multi-core) machine. During execution, however, only 1 core is utilised, while the rest of the CPU cores are idle. The cores are under-utilised although I have configured the MonetDB server to use all the physical cores of my machine (either by setting 'nthreads' or, in the case of mserver5, 'gdk_nr_threads'). I am running MonetDB server v11.21.13 "Jul2015-SP2", compiled from source. Is this due to some configuration setting I missed? Or, due to the fact that I am running a single-operator query? Alternatively, I was thinking of creating merge tables, thus partitioning my tables manually, hoping that this will allow for intra-operator parallelism. Thanks, Alexandros

2 1

create table uppercase with " "
by Pierre-Adrien Coustillas 20 Jan '16

20 Jan '16

Hello, bug or feature ? * PETER in uppercase : sql>CREATE TABLE "PETER" ( nom varchar(20) ); operation successful (2.545ms) sql>DROP TABLE PETER; DROP TABLE: no such table 'peter' sql>DROP TABLE "PETER"; operation successful (1.947ms) * peter in lowercase : sql>CREATE TABLE "peter" ( nom varchar(20) ); operation successful (2.917ms) sql>DROP TABLE peter; operation successful (1.882ms) * peter in lowercase and uppercase : sql>CREATE TABLE "peter" ( nom varchar(20) ); operation successful (2.900ms) sql>CREATE TABLE PETER ( nom varchar(20) ); CREATE TABLE: name 'peter' already in use sql>CREATE TABLE "PETER" ( nom varchar(20) ); operation successful (9.339ms) Conclusion : PETER is convert in peter "PETER" is not convert Pierre Coustillas -- 1G6 52 route de bischwiller 67300 Schiltigheim Société de Services et de Formations en Logiciels Libres http://1g6.biz Tél : 06 64 63 70 35 // 09 54 38 70 90

2 1

best practice use schema
by Pierre-Adrien Coustillas 20 Jan '16

20 Jan '16

Hello, What is the best practice on the use of schema ? S hould we always create a schema , or is can I use the " sys" schema by default ? Pierre Coustillas -- 1G6 52 route de bischwiller 67300 Schiltigheim Société de Services et de Formations en Logiciels Libres http://1g6.biz Tél : 06 64 63 70 35 // 09 54 38 70 90

1 0

creating a binary file for "COPY binary ..."
by Lynn Carol Johnson 19 Jan '16

19 Jan '16

Hi all - I'm creating a large db table with several columns. To speed it up, I'm using monetdb's "COPY BINARY" command. I have a question on creating the binary files. What I've done is use java's DataOutputStream to create the binary file - using code that looks as below (sometimes an int, sometimes a real or byte): writerChr = new DataOutputStream(new BufferedOutputStream(new FileOutputStream(chrFile))); writerChr.writeInt(intChrom); When I read this file using a test program and java's DataInputStream, it correctly shows me the integer values I've written. In the "writerChr" case, all values in this particular column are "10". However, when I load this into monetdb the value is way off - it shows a large number. The command I am using to copy this file and one other into a table with 2 "int" columns is: COPY binary into justchrpos from ('/home/lcj34/chrFileC10.bin','/home/lcj34/posFileC10.bin'); It copies in fine, but when I cheek the values I see this: (not 10!) sql>\d justchrpos CREATE TABLE "testbig"."justchrpos" ( "chr" INTEGER, "pos" INTEGER ); sql>select * from justchrpos limit 10; +-----------+-----------+ | chr | pos | +===========+===========+ | 167772160 | 16777216 | | 167772160 | 33554432 | | 167772160 | 50331648 | | 167772160 | 67108864 | | 167772160 | 83886080 | | 167772160 | 100663296 | | 167772160 | 117440512 | | 167772160 | 134217728 | | 167772160 | 150994944 | | 167772160 | 167772160 | +-----------+-----------+ 10 tuples (3.601ms) sql> QUESTION: Is using java's DataOutputStream a valid way to create the binary files for monetdb? If not, what should I be using? As I mentioned above, when I read back the created file using java's DataInputStream, I see the values "10" as I expect. Thanks - Lynn

3 3

Monet way of handling string columns
by bhuvan Mitra 18 Jan '16

18 Jan '16

Hello, I have recently started looking into Monetdb and want to contribute to it. My complete focus is on string columns as numeric data can already be processed with maximum speed. I know it uses BATs to store every single column, but I read some where that it uses dictionary for string columns. How is it actually using dictionary? And are there any other special handling techniques for strings? When a query like 'contains' comes, how will monet handle that? I found there is no inverted index in monet to deal queries like this. (Please tell a rough model of how query execution is done incase of string columns) Hope this forum gives me some insight of what's happening during string column processing. Thanks in advance, -Mitra

1 0

MonetDB-11.21.13 (Jul2015-SP2) can't start in CentOS 6.3
by chenkaijiang＠bigknow.com.cn 18 Jan '16

18 Jan '16

I downloaded the source codes at https://www.monetdb.org/downloads/sources/Jul2015-SP2/MonetDB-11.21… ./configure --prefix=/usr/local/MonetDBmakemake install No error is reported. But can't start:monetdbd create ./ycmonetdbd: error while loading shared libraries: libstream.so.7: cannot open shared object file: No such file or directory It looks like the libstream.so is lib on debian linux.  How can I fix it on CentOS 6.3?

2 1

How to set query timeout in MonetDB Server
by Luciano Sasso 16 Jan '16

16 Jan '16

How to set query timeout in MonetDB Server? -- Luciano Sasso Vieira Data Scientist & Solutions Architect luciano(a)gsgroup.com.br <http://www.gsgroup.com.br> | tel: 17 3353-0833 | cel: 17 99706-9335 www.gsgroup.com.br <http://www.gsgroup.com.br> --- Este email foi escaneado pelo Avast antivírus. https://www.avast.com/antivirus

1 0