December 2010 - users-list

[MonetDB-users] Proofreading your professional documents
by Richard Hounslow 24 Feb '24

24 Feb '24

Dear Professor/Lecturer Please excuse us for contacting you directly. We are offering an efficient service to help you ensure your academic and professional work is written in perfect English. We will check the grammar and style of your work and return it to you to meet your requirements and deadlines. Proof-Reading-Service.com can provide you with a professional proofreading service at a very reasonable rate. All our proofreaders are highly qualified native English speakers. Many work as leading academics in their fields and all have extensive experience of proofreading to the highest standards. If you are interested in our service, please take a look at our website: www.proof-reading-service.com All you have to do is send us your document as a word attachment with the deadline and we will guarantee delivery of a perfectly written document to give you complete confidence when you submit your work. The fee is worked out on a flat rate (£6.95 per thousand words or 0.695 pence per word), so you know exactly how much the proofreading will cost in advance. Once we have received your document(s) we will confirm the word count, the price and the deadline. If you have not heard from us within 2 hours during normal business hours after you have sent your work please resend it from a different email account or send the file to proofreadings(a)gmail.com. We look forward to hearing from you! Yours faithfully Proof-Reading-Service.com

2 2

[MonetDB-users] Add primary key performance
by paulr135 13 Jan '11

13 Jan '11

I've loaded a 27GB (~165M rows) file into a table in MonetDB and now I am trying to add a primary key which consists of 4 fields - 2 varchars and 2 ints. It has been running for a few days (since Friday of last week). I've followed the instructions laid out on this post: http://old.nabble.com/MonetDB-SQL:-bulk-loading-large-data-sets---adding-ke… which involves setting the optimizer variable to "nov2009_pipe". I am unable to log in via mclient - I am able to put in my credentials but it just hangs at the "Welcome to mclient, the MonetDB/SQL interactive terminal (Jun2010-SP2)" line. So I have a few questions: 1. Is there any way I can see the status of the alter table ... add primary key command? 2. Is it possible to stop the alter table gracefully? Or will I have to shut the db down? 3. Is there anything I should do in future to speed up adding of primary keys? Thanks. -- View this message in context: http://old.nabble.com/Add-primary-key-performance-tp30397515p30397515.html Sent from the monetdb-users mailing list archive at Nabble.com.

2 3

[MonetDB-users] Monnetdb/Xquery
by Ramachandran, Prakash 30 Dec '10

30 Dec '10

Hi, I am currently evaluating software's for metadata storage that can be used in our product. I came across your product and found it very interesting and we might consider using it for our metadata storage. I just wanted to know if Monnetdb provides commercial support, if so can you please provide the detail on sales contacts. Thank in advance. Regards, Prakash R

1 0

[MonetDB-users] MAPI interface error
by Brett Spurrier 24 Dec '10

24 Dec '10

Hi all, I am having a bit of trouble and I am hoping a seasoned MonetDB expert can lend some advice. I am using the MAPI interface in my C/C++ application to talk to my MonetDB instance. I have a recursive INSERT INTO statement which populates the database, and it initially works pretty well. However, as I populate more (past 200 inserts in my case), I get a MAPI error (exampled by: mapi_explain_result()): MAPI = monetdb@localhost:50000 ACTION= mapi_start_talking ERROR = !Connection terminated No real reason as to why. Has anyone seen this before? My code: MapiHdl query(Mapi dbh, char *q) { MapiHdl ret = NULL; if ((ret = mapi_query(dbh, q)) == NULL || mapi_error(dbh) != MOK) die(dbh, ret); return(ret); } void update(Mapi dbh, char *q) { MapiHdl ret = query(dbh, q); if (mapi_close_handle(ret) != MOK) die(dbh, ret); } void main() { dbh = mapi_connect("localhost", 50000, "monetdb", "monetdb", "sql", "dbname"); if (mapi_error(dbh)) { die(dbh, hdl); } for (int i=0; i< 2500; i++) { char *p = INSERT_INTO_STATEMENT; update(dbh, p); } } Does anyone see any reason that the MAPI would choke mid way through the For loop? Any help would be majorly appreciated. I am at a loss. Cheers, Brett

3 14

[MonetDB-users] Announcement: New Jun2010-SP1 bugfix release of MonetDB suite
by Sjoerd Mullender 21 Dec '10

21 Dec '10

The MonetDB team at CWI/MonetDB BV is pleased to announce the Oct2010-SP1 bug fix release of the MonetDB suite of programs. The most important and visible change is a security fix in the SQL COPY INTO command. This is detailed in the release notes. More information (including release notes) on this release is available at <http://monetdb.cwi.nl/Development/Releases/Oct2010/>. The download location is <http://dev.monetdb.org/downloads/>. -- Sjoerd Mullender

1 0

[MonetDB-users] How to limit monetdb memory consumption on Windows
by Bianco Prieto-Ilarion 18 Dec '10

18 Dec '10

Hi all! I have a problem. I'm using Oct2010 64-bit MonetDB build. I have a table with 8M records and when I'm trying to execute LEFT JOIN (from JDBC) MonetDB is crashed. I can see that it crashed due to memory (my laptop has 8Gb) consumption is unlimited. What can you suggest (just use Linux is a final step:)) Thanks, Bina Prieto

1 0

[MonetDB-users] MonetDB Installation
by Rob Fynes 16 Dec '10

16 Dec '10

Hello, Regarding the pre-compiled binaries directories for deb distributions "karmic" and "lucid" in the downloads section, there doesn't appear to be anything for x-query. Is this intentional? Regards, -Rob

2 2

[MonetDB-users] mserver5 ignores ulimit?
by Siva Manoharan 14 Dec '10

14 Dec '10

Hi. I am currently running monetdb (release Oct2010) on Debian etch, with 8GB physical RAM and 10GB virutal memory. During unrestricted operation (using the merovingian daemon) monetdb takes up a significant majority of the resources, which is not desirable as there are other processes running on the same server. To limit the memory, I tried using ulimit (setting -d, -v and -m all to 2GB). According to the logs, merovingian notices the limit (a line from log states "Found 1.907 GiB available main-memory"). However, memory usage of mserver still exceeds that limit and is killed off by the OS, abandoning an ongoing query. (merovingian is not killed and mserver restarts without any problems). Is this expected behaviour or is it a bug? Is there another way to limit monetdb's memory usage? As a note, none of the tables (and therefore none of the columns) are anywhere near 2GB in size. The biggest table has ~2.5M rows, where each row has 2 columns ( 8byte bigint and 4 byte int). However, there are thousands of tables, and mserver exceeds the limit most often when a large number of tables are being updated within a short time (in a single php process, however). Is it possible that mserver doesn't release memory fast enough to allow multiple tables to be accessed? Thanks.

1 0

[MonetDB-users] crashes with copy into (270MB file)
by Lex Slaghuis 13 Dec '10

13 Dec '10

Hello I Experience crashes with the copy into statement, with large files (270MB utf8) ... sql>COPY 6000000 records into wptool.wpstats from '/var/www/pagecounts' using delimiters ' ','\n','' null as ''; MAPI = testuser@localhost:50000 ACTION= read_line QUERY = COPY 6000000 records into wptool.wpstats from '/var/www/pagecounts' using delimiters ' ','\n','' null as ''; ERROR = !Connection terminated Timer 5846.447 msec...... Number of records: wc -l /var/www/pagecounts 5983201 /var/www/pagecounts My memory: free -m(egabytes) total used free shared buffers cached Mem: 2007 431 1576 0 20 304 -/+ buffers/cache: 106 1900 Swap: 16769 0 16769 Error-logs: 2010-12-13 14:17:41 MSG merovingian[902]: proxying client localhost:52737 for database 'test' to mapi:monetdb:///var/MonetDB5/dbf arm/test/.mapi.sock?database=test 2010-12-13 14:17:41 MSG merovingian[902]: target connection is on local UNIX domain socket, passing on filedescriptor instead of proxying 2010-12-13 14:18:30 MSG merovingian[902]: database 'test' (1142) was killed by signal SIGSEGV 2010-12-13 14:18:38 ERR control[902]: (local): received stop signal for non running database: test 2010-12-13 14:18:46 MSG merovingian[902]: database 'test' has crashed after start on 2010-12-13 14:17:33, attempting restart, up min/avg/max: 14s/22h/2d, crash average: 1.00 0.70 0.23 (10-3=7) Some specs: uname -a Linux ubuntu 2.6.32-24-server #43-Ubuntu SMP Thu Sep 16 16:05:42 UTC 2010 x86_64 GNU/Linux 2010-12-13 14:17:33 MSG test[1142]: # Compiled for x86_64-pc-linux-gnu/64bit with 64bit OIDs dynamically linked 2010-12-13 14:17:33 MSG test[1142]: # Found 1.961 GiB available main-memory. If anyone has some pointers on this , it would really help out. I think it would be really to chop up the files into smaller pieces, just for import. Regards Lex Slaghuis -----Original Message----- From: monetdb-users-request(a)lists.sourceforge.net [mailto:monetdb-users-request@lists.sourceforge.net] Sent: vrijdag 10 december 2010 11:12 To: monetdb-users(a)lists.sourceforge.net Subject: MonetDB-users Digest, Vol 55, Issue 2 Send MonetDB-users mailing list submissions to monetdb-users(a)lists.sourceforge.net To subscribe or unsubscribe via the World Wide Web, visit https://lists.sourceforge.net/lists/listinfo/monetdb-users or, via email, send a message with subject or body 'help' to monetdb-users-request(a)lists.sourceforge.net You can reach the person managing the list at monetdb-users-owner(a)lists.sourceforge.net When replying, please edit your Subject line so it is more specific than "Re: Contents of MonetDB-users digest..." Today's Topics: 1. COPY INTO: wrong length with backslashed chars? (Daniel Boesswetter) 2. Re: COPY INTO: wrong length with backslashed chars? (Stefan de Konink) 3. Some questions regarding parallelism and other stuff... (Mike De-La-Columnar) 4. Re: Some questions regarding parallelism and other stuff... (Martin Kersten) 5. Add primary key performance (paulr135) 6. Re: Add primary key performance (Stefan Manegold) 7. Re: COPY INTO: wrong length with backslashed chars? (Daniel Boesswetter) ---------------------------------------------------------------------- Message: 1 Date: Sun, 05 Dec 2010 19:08:01 +0100 From: Daniel Boesswetter <daniel.boesswetter(a)fu-berlin.de> Subject: [MonetDB-users] COPY INTO: wrong length with backslashed chars? To: monetdb-users(a)lists.sourceforge.net Message-ID: <4CFBD501.9030709(a)fu-berlin.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Hi all, I'm using MonetDB for the analysis for ~90GB of data which come from a MySQL-based web-application. The data is dumped into CSV-files on a regular basis and then imported into MonetDB with the COPY command. I have experienced some strange behaviour: obviously, MonetDB counts the number of chars for a VARCHAR-field before removing backslashes which are used e.g. for escaping newlines or other backslashes and rejects strings that should actually fit. A simple reproduction is this: 1. In mclient: CREATE TABLE t (x VARCHAR(1)); 2. Create a textfile foo.txt containing two backslashes: \\ 3. In mclient: sql>copy into t from '/tmp/foo.txt'; SQLException:importTable:value '\\' from line 1 field 1 not inserted, expecting type varchar(1) failed to import table Performing exactly the same steps with a VARCHAR(2) column works without errors and results in t.x containg the de-escaped single backslash. I'm using the Oct-2010 release. My current workaround is to enlarge VARCHARs which caused problems in the past, but this is of course no guarantee that it won't happen again. Although I found a few monetdb-bugs entries concerning the COPY command (e.g. about the misleading error message seen above), none seems to deal with my problem. Is this a bug? Best regards, Daniel ------------------------------ Message: 2 Date: Sun, 05 Dec 2010 19:15:36 +0100 From: Stefan de Konink <stefan(a)konink.de> Subject: Re: [MonetDB-users] COPY INTO: wrong length with backslashed chars? To: Communication channel for MonetDB users <monetdb-users(a)lists.sourceforge.net> Message-ID: <4CFBD6C8.30208(a)konink.de> Content-Type: text/plain; charset=ISO-8859-1 -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Op 05-12-10 19:08, Daniel Boesswetter schreef: > Hi all, > > I'm using MonetDB for the analysis for ~90GB of data which come from a > MySQL-based web-application. The data is dumped into CSV-files on a > regular basis and then imported into MonetDB with the COPY command. I > have experienced some strange behaviour: obviously, MonetDB counts the > number of chars for a VARCHAR-field before removing backslashes which > are used e.g. for escaping newlines or other backslashes and rejects > strings that should actually fit. > > A simple reproduction is this: > > 1. In mclient: CREATE TABLE t (x VARCHAR(1)); 2. Create a textfile > foo.txt containing two backslashes: \\ 3. In mclient: > sql>copy into t from '/tmp/foo.txt'; > SQLException:importTable:value '\\' from line 1 field 1 not inserted, > expecting type varchar(1) failed to import table > > Performing exactly the same steps with a VARCHAR(2) column works > without errors and results in t.x containg the de-escaped single backslash. > > I'm using the Oct-2010 release. My current workaround is to enlarge > VARCHARs which caused problems in the past, but this is of course no > guarantee that it won't happen again. > > Although I found a few monetdb-bugs entries concerning the COPY > command (e.g. about the misleading error message seen above), none > seems to deal with my problem. > > Is this a bug? What are you escaping? Paste the line that fails ;) Stefan -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEAREKAAYFAkz71sgACgkQYH1+F2Rqwn3AmgCfayhLg6zjXZgvqhV9OHJbj5Kh z64An2xZJy7H1wv3M6hr/SHQaK2C8IM3 =//V/ -----END PGP SIGNATURE----- ------------------------------ Message: 3 Date: Mon, 6 Dec 2010 05:07:46 -0800 (PST) From: Mike De-La-Columnar <mikevermondusa(a)gmail.com> Subject: [MonetDB-users] Some questions regarding parallelism and other stuff... To: monetdb-users(a)lists.sourceforge.net Message-ID: <30366715.post(a)talk.nabble.com> Content-Type: text/plain; charset=us-ascii Hi fellas, If its not too much of a hassle, i have a few questions regarding parallelism & the internal algorithms 1. What is the difference between Data-Flow, and Mitosis? 2. Is there a special reason why the distinct count aggregation is not parallelized just like the other aggregation functions? Is it just a matter of not making it into the release, or is there a fundamental problem here. Note: even though the parallel distinct can be slower in certain distributions, it could still be useful... 3. Are there any plans on supporting parallel frameworks like OpenMP/Intel TBB directly within the low-level MAL functions? Providing data-parallelism on the loop-code level could even increase performance. Or did you already play with it and decided that the performance gain is just not worth it? 4. Is there a way for me to prevent MonetDB server from cleaning up all intermediate BATs between server restarts. I've read here that some intermediates (like join indices etc.) are lost between restarts... is it still correct? Thank you very much for your time, Keep up the amazing job of building the most powerful columnar database out there! -- View this message in context: http://old.nabble.com/Some-questions-regarding-parallelism-and-other-stuff.… Sent from the monetdb-users mailing list archive at Nabble.com. ------------------------------ Message: 4 Date: Mon, 06 Dec 2010 20:13:55 +0100 From: Martin Kersten <Martin.Kersten(a)cwi.nl> Subject: Re: [MonetDB-users] Some questions regarding parallelism and other stuff... To: monetdb-users(a)lists.sourceforge.net Message-ID: <4CFD35F3.1060001(a)cwi.nl> Content-Type: text/plain; charset=ISO-8859-1; format=flowed On 12/6/10 2:07 PM, Mike De-La-Columnar wrote: > > Hi fellas, Hi Mike > > If its not too much of a hassle, i have a few questions regarding > parallelism& the internal algorithms > > 1. > What is the difference between Data-Flow, and Mitosis? See optimizer descriptions: http://monetdb.cwi.nl/MonetDB/Documentation/Optimizer-Toolkit.html Orthogonal issues. > > 2. > Is there a special reason why the distinct count aggregation is not > parallelized just like the other aggregation functions? > Is it just a matter of not making it into the release, or is there a > fundamental problem here. Distinct is a global property, not a local one. A distributed version may easily lead to throwing the complete underlying table around. > > Note: even though the parallel distinct can be slower in certain > distributions, it could still be useful... > > 3. > Are there any plans on supporting parallel frameworks like OpenMP/Intel TBB > directly within the low-level MAL functions? > Providing data-parallelism on the loop-code level could even increase > performance. It could improve, but there should be a clear driving application. > Or did you already play with it and decided that the performance gain is > just not worth it? Depends on the amount of work per byte unit. > > 4. > Is there a way for me to prevent MonetDB server from cleaning up all > intermediate BATs between server restarts. > I've read here that some intermediates (like join indices etc.) are lost > between restarts... is it still correct? All temporaries are indeed lost. Join indices for foreign key checking are retained on disk. Rebuilding hashes is cheaper (mostly) then write/read them from disk > > > Thank you very much for your time, > Keep up the amazing job of building the most powerful columnar database out > there! Thanks for the support !. Martin > > > ------------------------------ Message: 5 Date: Tue, 7 Dec 2010 07:44:14 -0800 (PST) From: paulr135 <romesh135(a)gmail.com> Subject: [MonetDB-users] Add primary key performance To: monetdb-users(a)lists.sourceforge.net Message-ID: <30397515.post(a)talk.nabble.com> Content-Type: text/plain; charset=us-ascii I've loaded a 27GB (~165M rows) file into a table in MonetDB and now I am trying to add a primary key which consists of 4 fields - 2 varchars and 2 ints. It has been running for a few days (since Friday of last week). I've followed the instructions laid out on this post: http://old.nabble.com/MonetDB-SQL:-bulk-loading-large-data-sets---adding-ke… which involves setting the optimizer variable to "nov2009_pipe". I am unable to log in via mclient - I am able to put in my credentials but it just hangs at the "Welcome to mclient, the MonetDB/SQL interactive terminal (Jun2010-SP2)" line. So I have a few questions: 1. Is there any way I can see the status of the alter table ... add primary key command? 2. Is it possible to stop the alter table gracefully? Or will I have to shut the db down? 3. Is there anything I should do in future to speed up adding of primary keys? Thanks. -- View this message in context: http://old.nabble.com/Add-primary-key-performance-tp30397515p30397515.html Sent from the monetdb-users mailing list archive at Nabble.com. ------------------------------ Message: 6 Date: Wed, 8 Dec 2010 09:38:23 +0100 From: Stefan Manegold <Stefan.Manegold(a)cwi.nl> Subject: Re: [MonetDB-users] Add primary key performance To: Communication channel for MonetDB users <monetdb-users(a)lists.sourceforge.net> Message-ID: <20101208083823.GA24931(a)cwi.nl> Content-Type: text/plain; charset=us-ascii On Tue, Dec 07, 2010 at 07:44:14AM -0800, paulr135 wrote: > > I've loaded a 27GB (~165M rows) file into a table in MonetDB and now I am > trying to add a primary key which consists of 4 fields - 2 varchars and 2 > ints. > It has been running for a few days (since Friday of last week). > > I've followed the instructions laid out on this post: > http://old.nabble.com/MonetDB-SQL:-bulk-loading-large-data-sets---adding-ke… > which involves setting the optimizer variable to "nov2009_pipe". Using the nov2009_pipe optimizer pipeline for adding constraint only holds for the Feb2010 release of MonetDB. It should not be required anymore for later releases as of Jun2010-SP1, incl. the latest Oct2010 release. Which version of MonetDB are you using? Which OS are you running on? What kind of manchine/hardware are you running on, in particular how much memory does it have? Can you check what the machine is doing? Is it running out of physical memory and swapping, i.e., is the virtual size of the mserver5 (much) larger than the machine's physical memory size? > I am unable to log in via mclient - I am able to put in my credentials but > it just hangs at the "Welcome to mclient, the MonetDB/SQL interactive > terminal (Jun2010-SP2)" line. > > So I have a few questions: > 1. Is there any way I can see the status of the alter table ... add primary > key command? You can use MonetDB's stethoscope tool to "listen" what the server is doing; however, while it is busy with a single (MAL-)operation, there will not be any output; hence, you might have to wait for some time to see some output. Unless very familiar with MonetDB internals, the output of stethoscope might not be too informative for you. But you're welcome to send us the output to have a look at it. In addition, you can check what the mserver5 is doing as suggested above. > 2. Is it possible to stop the alter table gracefully? Or will I have to > shut the db down? No, unfortunately there is no way to gracefully stop a running query. You'd have to way until it finishes (successfully or with an error), or shut down the DB. > 3. Is there anything I should do in future to speed up adding of primary > keys? In priciple not. We are aware that with data sizes (i.e., total size of all columns that the primary keys consists of) that significantly exceed main memory, adding the key (i.e., checking whether the data fulfills the key condition) does not work as efficiently as it could/should, in particular with multi-column keys --- we're working on that ... You could, though, run the ALTER TABLE statement in trace mode, i.e., `TRACE ALTER TABLE ...`, collect the output and send it to us for analysis; that could help us to improve the creation of large multi-column primary keys. Regards, Stefan > Thanks. > -- > View this message in context: http://old.nabble.com/Add-primary-key-performance-tp30397515p30397515.html > Sent from the monetdb-users mailing list archive at Nabble.com. > > > ------------------------------------------------------------------------------ > What happens now with your Lotus Notes apps - do you make another costly > upgrade, or settle for being marooned without product support? Time to move > off Lotus Notes and onto the cloud with Force.com, apps are easier to build, > use, and manage than apps on traditional platforms. Sign up for the Lotus > Notes Migration Kit to learn more. http://p.sf.net/sfu/salesforce-d2d > _______________________________________________ > MonetDB-users mailing list > MonetDB-users(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/monetdb-users > > > > > > > > > > > -- | Stefan.Manegold @ CWI.nl | DB Architectures (INS1) | | http://CWI.nl/~manegold/ | Science Park 123 (L321) | | Tel.: +31 (0)20 592-4212 | 1098 XG Amsterdam (NL) | ------------------------------ Message: 7 Date: Fri, 10 Dec 2010 11:11:34 +0100 From: Daniel Boesswetter <daniel.boesswetter(a)fu-berlin.de> Subject: Re: [MonetDB-users] COPY INTO: wrong length with backslashed chars? To: monetdb-users(a)lists.sourceforge.net Message-ID: <4D01FCD6.9040201(a)fu-berlin.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Hi Stefan, > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA512 > > Op 05-12-10 19:08, Daniel Boesswetter schreef: > > Hi all, > > > > I'm using MonetDB for the analysis for ~90GB of data which come from a > > MySQL-based web-application. The data is dumped into CSV-files on a > > regular basis and then imported into MonetDB with the COPY command. I > > have experienced some strange behaviour: obviously, MonetDB counts the > > number of chars for a VARCHAR-field before removing backslashes which > > are used e.g. for escaping newlines or other backslashes and rejects > > strings that should actually fit. > > > > A simple reproduction is this: > > > > 1. In mclient: CREATE TABLE t (x VARCHAR(1)); > > 2. Create a textfile foo.txt containing two backslashes: \\ > > 3. In mclient: > > sql>copy into t from '/tmp/foo.txt'; > > SQLException:importTable:value '\\' from line 1 field 1 not inserted, > > expecting type varchar(1) > > failed to import table > > > > Performing exactly the same steps with a VARCHAR(2) column works without > > errors and results in t.x containg the de-escaped single backslash. > > > > I'm using the Oct-2010 release. My current workaround is to enlarge > > VARCHARs which caused problems in the past, but this is of course no > > guarantee that it won't happen again. > > > > Although I found a few monetdb-bugs entries concerning the COPY command > > (e.g. about the misleading error message seen above), none seems to deal > > with my problem. > > > > Is this a bug? > > What are you escaping? Paste the line that fails ;) > > As I said: try to import a single "\" into a VARCHAR(1) with COPY. Writing "\" does not work because it escapes the newline. Writing "\\" does not work because it is longer than 1 char. Regards, Daniel > Stefan > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v2.0.16 (GNU/Linux) > Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ > > iEYEAREKAAYFAkz71sgACgkQYH1+F2Rqwn3AmgCfayhLg6zjXZgvqhV9OHJbj5Kh > z64An2xZJy7H1wv3M6hr/SHQaK2C8IM3 > =//V/ > -----END PGP SIGNATURE----- > ------------------------------ ------------------------------------------------------------------------------ ------------------------------ _______________________________________________ MonetDB-users mailing list MonetDB-users(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/monetdb-users End of MonetDB-users Digest, Vol 55, Issue 2 ********************************************

1 0

[MonetDB-users] COPY INTO: wrong length with backslashed chars?
by Daniel Boesswetter 10 Dec '10

10 Dec '10

Hi all, I'm using MonetDB for the analysis for ~90GB of data which come from a MySQL-based web-application. The data is dumped into CSV-files on a regular basis and then imported into MonetDB with the COPY command. I have experienced some strange behaviour: obviously, MonetDB counts the number of chars for a VARCHAR-field before removing backslashes which are used e.g. for escaping newlines or other backslashes and rejects strings that should actually fit. A simple reproduction is this: 1. In mclient: CREATE TABLE t (x VARCHAR(1)); 2. Create a textfile foo.txt containing two backslashes: \\ 3. In mclient: sql>copy into t from '/tmp/foo.txt'; SQLException:importTable:value '\\' from line 1 field 1 not inserted, expecting type varchar(1) failed to import table Performing exactly the same steps with a VARCHAR(2) column works without errors and results in t.x containg the de-escaped single backslash. I'm using the Oct-2010 release. My current workaround is to enlarge VARCHARs which caused problems in the past, but this is of course no guarantee that it won't happen again. Although I found a few monetdb-bugs entries concerning the COPY command (e.g. about the misleading error message seen above), none seems to deal with my problem. Is this a bug? Best regards, Daniel

3 4