October 2015 - users-list

Presto Connector for MonetDB
by Stephen P. Morgan 20 Oct '15

20 Oct '15

Does anyone know of a Presto connector for MonetDB? There are connectors for MySQL and PostgreSQL among others, but I don't see one for MonetDB. Thanks, Steve

2 1

Transactions
by Doug Service 19 Oct '15

19 Oct '15

We use MonetDB extensively in an internal system we are developing. We initially had issues getting transactions working so we put it off to make progress in other areas. However we are at the point now were we need to have transactions working. The mapi documentation ( https://www.monetdb.org/Documentation/Manuals/SQLreference/Programming/MAPI) says the following about auto commit. MapiMsg mapi_setAutocommit(Mapi mid, int autocommit) Set the autocommit flag (default is on). This only has an effect when the language is SQL. In that case, the server commits after each statement sent to the server. I infered from this statement that using transactions requires setting auto commit to false. However, when I did that I received the following error when I tried to use transactions. MSERVER(25001!START TRANSACTION: cannot start a transaction within a transaction) If I did not disable auto commit, I did not receive any errors when using transactions, but I did not know if the transaction were actually working. After using hg to enlist in the MonetDb code and searching for the error I found this. str SQLtransaction2(Client cntxt, MalBlkPtr mb, MalStkPtr stk, InstrPtr pci) { mvc *sql = NULL; str msg; (void) stk; (void) pci; if ((msg = getSQLContext(cntxt, mb, &sql, NULL)) != NULL) return msg; if ((msg = checkSQLContext(cntxt)) != NULL) return msg; if (sql->session->auto_commit == 0) throw(SQL, "sql.trans", "25001!START TRANSACTION: cannot start a transaction within a transaction"); if (sql->session->active) { RECYCLEdrop(cntxt); mvc_rollback(sql, 0, NULL); } sql->session->auto_commit = 0; sql->session->ac_on_commit = 1; sql->session->level = 0; (void) mvc_trans(sql); return msg; } which leads me to believe that one should not disable auto commit to use transaction as it appears to be auto disabled by the transaction code. My questions are: 1) Is the correct method of using transactions in MonetDb to leave the auto commit flag alone and just issue START TRANSACTION;, COMMIT;, and ROLLBACK; as needed? 2) If the answer to (1) is yes, is disabling auto commit, issuing updates, and then re enabling auto commit like a transaction? Doug Service

3 3

MonetDB not cleaning intermediate results which leads to filling up disk space
by abraun_75＠yahoo.de 19 Oct '15

19 Oct '15

Hi, Right now I am evaluating MonetDB and try to analyze a dataset that consists of 10 million rows. The data set has 2 tables, table a stores 10 million rows, table b stores 1 million rows. Both tables are being joined and some statistical values are being calculated (QUARTILE, AVG, SUM and GROUP BY). MonetDB is promising fast here. When starting up the database and executing the query the result is being returned within 4 seconds. I tried to execute the same SQL query 1000 times sequentially and to my surprise after the 60th iteration monetDB crashed because of insuffcient disk space. As far as I read intermediate results are being materialized within the bat folder. The .tail files are not being removed after a query was executed successfully and the files are not even removed after the database conenction is being closed. When I restart the server (dbfarm) the files are cleaned up perfectly. I am just wondering if this default behaviour can be changed? Are there any ways to enforce cleaning up the files? How to avoid a server crash because of insufficient memory? For the system I am planning to use monetDB for I am not able to restart the server before I execute the SQL queries. Thank you very much, Alex

5 8

current schema
by Enzo 19 Oct '15

19 Oct '15

Thanks to Jennie I understand that I can get the current schema with the query: select current_schema from sys.var(); While I get the “right” response, I get it as a series of duplicated lines (10 at the last count): For example, after setting a schema as set schema viewer; I get the following 10 lines from the query: Is it the right behaviour? And if yes, do these lines have a specific significance? Thanks Enzo enzo(a)smartinsightsfromdata.com

2 1

Re: users-list Digest, Vol 38, Issue 9
by Luciano Sasso 18 Oct '15

18 Oct '15

Jul 2015 is avaible for DEbian 7 ? At --- Este email foi escaneado pelo Avast antivírus. https://www.avast.com/antivirus

3 2

How to install tachograph?
by Xu,Wenjian 18 Oct '15

18 Oct '15

Hi all, I have installed the latest Jul2015 release, and I would like to use the latest tools tachograph to determine the progress of query execution. But I do not know how to install it. My system is ubuntu 14.04. I have tried to use *sudo apt-get install monetdb-tools* but failed to locate the package *monetdb-tools*. Any idea would be helpful. Thanks. Best, XU Wenjian

3 2

DataCell Stream
by Luciano Sasso 18 Oct '15

18 Oct '15

DataCel stream discontinead in new versions MonetDB ? A Idea, substitute for Proxy Node.JS ? --- Este email foi escaneado pelo Avast antivírus. https://www.avast.com/antivirus

3 3

Re: users-list Digest, Vol 38, Issue 9
by Luciano Sasso 18 Oct '15

18 Oct '15

Im Jul 2015, remote tables can sort in result query ? At --- Este email foi escaneado pelo Avast antivírus. https://www.avast.com/antivirus

2 1

MonetDB scripting
by Doug Service 18 Oct '15

18 Oct '15

Hello Everyone, I need to become more conversant in monetdb scripting and have read the pages on functions and procedures on the website. However, more examples would be very helpful. Does anyone know of a good repository of extensive monetdb scripts that provide good models. Doug Service

2 1

Concurrent Transaction control in MonetDB
by Vijay Krishna 18 Oct '15

18 Oct '15

Hi, I read the following lines in a thesis on MonetDB titled - "Improving Transactional Stability in MonetDB" where in they had explained the concurrency control in MonetDB as "Concurrent updates (writers) are not anticipated and in case of N con􏰃icting transactions, only one of them (the one having the earliest time-stamp) is guaranteed to eventually commit its changes. Concurrent appends (on the same table) are detected as write-to-write conflicts and only the writer carrying the earliest time-stamp will end up committing successfully, the rest will be aborted." Why is that the transaction having the earliest timestamp is allowed to commit, in case of a conflict? Shouldn't it be the latest transaction that should commit on a data, since the latest transaction would be having the latest version of data? Or, what am I getting wrong here? Any help much appreciated. Thanks & Regards, Vijayakrishna.P. Mobile : (+91) 9500402305.

1 0