DBfarm1

I will login as admin into "voc" database.

monetdbd start /home/fffovde/DBfarm1 mclient -u monetdb -d voc     #password monetdb

System Procedures

Table Statistics

In one of the previous blog posts "link", we saw the existence of a system table containing statistics about database tables:

SELECT * FROM sys.statistics WHERE minval = 'A';

We saw that this table can be updated partially with these commands. Now, we will see that for each of these commands we can use stored procedure, as an alternative.

	Update schema statistics:	Update table statistics:	Update column statistics:
Command	ANALYZE sys;	ANALYZE sys.auths;	ANALYZE sys.auths ( name );
Procedure	CALL sys.analyze('sys');	CALL sys.analyze('sys', 'auths');	CALL sys.analyze( 'sys', 'auths', 'name' );

There is also a procedure that will update statistics for all the schemas:

CALL sys.analyze();

Table Data Dump

I will read from one system table. We'll see that this table is empty.   SELECT * FROM sys.dump_statements;

I will now call this procedure: CALL sys.dump_table_data( 'sys', 'auths' );

We will read again from the "sys.dumpstatements".   SELECT * FROM sys.dump_statements; Now we have a row with a statement that can be used to fill "auths" table with the data.

We can empty this table normally: DELETE FROM sys.dump_statements;

If we call the same procedure, but without arguments, all the tables from the database will be dumped ( image is from "DBfarmDB" database ):   CALL sys.dump_table_data(); SELECT * FROM sys.dump_statements;

Eval Procedure

I want to find the table in my database that has the biggest ID. I want to read the content of that table. It is easy to find the name of this table. The problem is how to read its content. SELECT name FROM sys.tables WHERE id = ( SELECT MAX( id ) FROM sys.tables );

CREATE OR REPLACE PROCEDURE TableWithBiggestID() BEGIN     DECLARE TableName VARCHAR(50);           SET TableName = ( SELECT name FROM sys.tables WHERE id = ( SELECT MAX( id ) FROM sys.tables ) );     CALL sys.eval( 'SELECT * FROM ' || TableName || ';' ); END;

For that task, I will use stored procedure "sys.eval". It will transform text into statement.

In our procedure, we will find the name of the table with the biggest ID ( table "tabula" ). We will create string "SELECT * FROM tabula;". To execute string, we will transform it into statement with "sys.eval". If we call our procedure "TableWithBiggestID", we will see the content of the table "tabula". CALL TableWithBiggestID();

This function is only available to administrator.

Vacuum

Vacuuming is a maintenance process where we clean databases of obsolete data and defragment columns. After deleting or updating many rows, old records will not be deleted but will only be marked as deleted. This can lead to excessive disk usage and prevent sequential reading of column values.

We can manually vacuum one table or one column.

CALL sys.vacuum('voc', 'voyages');

CALL sys.vacuum('voc', 'voyages', 'number');

It is possible to schedule column vacuuming at regular intervals. The interval is expressed in seconds.

CALL sys.vacuum('voc', 'voyages', 'number', 3600);   This is how we stop the schedule.                                                                                                      CALL sys.stop_vacuum('voc', 'voyages','number');

Session Procedures

Session Procedures for Users

User can read his own session ID. SELECT sys.current_sessionid();

User can limit memory consumption for his queries. This is measured in MB. CALL sys.setmemorylimit(500);

User can disable his memory limit if he call this procedure with the argument equal to zero.   CALL sys.setmemorylimit(0);

Users can set query timeout to 1 second.   CALL sys.setquerytimeout(1);

If our query is longer than 1 second, error message will appear.

Zero will reset query timeout. CALL sys.setquerytimeout(0);

We can limit session to 3 seconds.   CALL sys.setsessiontimeout(3);

This will happen if we run a query after 3 seconds.

Zero will reset session timeout. CALL sys.setsessiontimeout(0);

Users can choose their optimizer. CALL sys.setoptimizer('minimal_pipe');

We can find a list of optimizers in the system table "sys.optimizers".

Default optimizer is "default_pipe".

We can limit the number of threads that user queries can use. CALL sys.setworkerlimit(8);

Zero argument will remove this limit. CALL sys.setworkerlimit(0);

Client Info Procedure for Users

We can declare some properties for the database client we are using. If we read from this system table, we will find a list of the properties that we can declare. SELECT * FROM sys.clientinfo_properties;

ClientHostname – the name of our computer. ApplicationName – Mclient, DBeaver, Tableau. ClientLibrary – libmapi 11.53.3 ClientPid – number of the client process. ClientRemark – A comment.

I will change one of these properties. CALL sys.setclientinfo('ClientHostname', 'OldServer' );

We can read current values of these properties in the "sys.sessions". SELECT * FROM sys.sessions;

The purpose is to write client data to the "sys.session" table, so we can better understand server usage.

Session Procedures for Admins

I will open another tab in the terminal, and there I will login as a user "voc".	mclient -u voc -d voc         --password is voc
If the user reads from "sys.session", he will only get his own session. SELECT * FROM sys.sessions;

When the admin reads from the "sys.sessions", he will see sessions with all the users. Admin can use IDs of those sessions to manage those sessions. Below we can see how admin can manage user voc session ( ID 1 ). SELECT * FROM sys.sessions;

CALL sys.setmemorylimit( 1, 100 );	Admin can limit memory usage to 100 MB.	We disable this limit by setting it to zero.
CALL sys.setoptimizer( 1, 'sequential_pipe');	Admin can set optimizer.	Default optimizer is "default_pipe".
CALL sys.setquerytimeout( 1, 8 );	User's queries will time out after 8 sec.	Zero will disable query time out.
CALL sys.setsessiontimeout( 1, 30 );	Session will time out after 30 seconds.	Zero will disable session time out.
CALL sys.setworkerlimit( 1, 8 );	User's queries are limited to 8 threads.	Zero will disable this limit.
CALL sys.stopsession( 1 );	We can stop user's session.

Queries Que

The Size of Queries Que

I will quit the session, and I will stop the "voc" database. quit monetdb stop voc

After that I will start the database, and I will login to it. monetdb start voc mclient -u monetdb -d voc    --password monetdb

I will read from the system view: select * from sys.queue; It is also possible to use function: select * from sys.queue();

The result is a table that has a history of the last run queries. This table was emptied when we restarted the database, so now it only has 2 queries.

I will create a list of 70 queries, and I will paste all of them in the mclient.

After all of them run, I will again read from the "sys.queue" table. SELECT COUNT(*) FROM sys.queue;

The result will be 64. MonetDB is keeping the last 64 queries.

If I login again as a "voc" user in the other tab of the terminal, I can run a query as a "voc" user.   mclient -u voc -d voc   --password voc SELECT 'New';

I will read "sys.queue" table as a "voc" user. SELECT * FROM sys.queue; He has 3 queries.

I will go back being admin, and I will again run the statement. SELECT COUNT( * ) FROM sys.queue();

The result will be 61. MonetDB is keeping only the last 64 queries ( 61 + 3 ), from the all of the sessions.

Admin has ability to read the whole que: select COUNT( * ) FROM sys.queue( 'ALL' );

With "ALL", admin can read from all of the sessions. SELECT DISTINCT sessionid FROM sys.queue( 'ALL' );

Changing The Que Size

The number of queries saved is the same as the maximal number of clients that can connect to database. I will quit my session. I will stop my database and then I will change "nclients" property of the database.

quit monetdb stop voc monetdb set nclients=10 voc

I will log in into database. mclient -u monetdb -d voc

I will again paste 70 queries into mclient.

If we now read from the "sys.queue", we will only find 10 queries. This is the proof that "nclients" property will limit number of queries kept inside of this table.

We can read the current value of "nclients" property with this query.   select * from environment where name = 'max_clients';

Management of Queries

We can provide the name of a user, and that will return only the queries for that user. Currently, all the saved queries belong to admin. select COUNT( * ) from sys.queue('monetdb');

"queue" view will also include queries that are currently running. Such queries will be labeled with "running" in the status column.

There are five possible statuses.

Query is executing. running

Query is halted. paused

Query is stopping. stopping

Query is stopped. aborted

Query is executed. finished

I will open one more tab in the terminal, where I will log in as a admin. mclient -u monetdb -d voc  --password monetdb

I will use this query on the right. This is recursive CTE that uses LEFT OUTER JOIN. Such query will never end. It is eternal. I will run this query in the first terminal tab, and I will check the result in the second terminal table.

WITH recursive numbers AS ( SELECT 1    UNION ALL   SELECT 1 FROM numbers        LEFT OUTER JOIN sys.auths ON 1 = 1 ) SELECT * FROM   numbers;

The last query in my "sys.queue" is number 93 ( column "tag" ). I know that when I run my eternal query, it will get number 94. I will run eternal query.

I will run this statement in the terminal second tab. My query will have status "running".

SELECT status FROM sys.queue WHERE tag = 94;

I can pause my query from the terminal second tab. Status will change.

CALL sys.pause( 94 );   SELECT status FROM sys.queue WHERE tag = 94;

I can resume my query. Status will change again.

CALL sys.resume( 94 );   SELECT status FROM sys.queue WHERE tag = 94;

We can stop the query.

CALL sys.stop( 94 );                                     SELECT status FROM sys.queue WHERE tag = 94;

It seems that my query can be only stopped with "Ctrl+C". This is because I used eternal query.

Even when we use "Ctrl+C" to stop query, its status will become "aborted".

SELECT status FROM sys.queue WHERE tag = 94;

If we are admin we can manage queries of other users. For that we must provide the name of a user and his query number ( from column "tag" ).

--CALL sys.pause( 999, 'User1');

--CALL sys.resume( 999, 'User1' );

--CALL sys.stop( 999, 'User1');

Mclient is the MonetDB command line client that connects to the MonetDB server to execute SQL commands. We have used this client in all the previous videos, but now we will see how we can make better use of it.

General Information

With mclient we can find out the version of our server and we can get help for mclient commands.   This will show us version:   mclient -v This is how we can get help: mclient -?

Information About Our Database

In one of the previous blogposts "link", I created database farm "DBfarm". Inside of it we have a database "DBfarmDB".   I will now login to that database as an admin and I will create some database objects. The objects will be created in the "sys" schema.   monetdbd start /home/fffovde/Desktop/DBfarm monetdb set readonly=no DBfarmDB            #if it is still readonly mclient -u monetdb -d DBfarmDB           #password monetdb

CREATE TABLE tablo( Number INT ); CREATE VIEW viewo( Number ) AS SELECT Number from tablo; CREATE SEQUENCE seqo;   CREATE FUNCTION funco() RETURNS INTEGER      RETURN SELECT 2;   CREATE SCHEMA schemo;

If we just type "\D" inside of the mclient, we will get sql dump in the stdout, for the whole database.

We can type the table name after the "\D" command. This will give us SQL to create that table.   \D tablo

If we want to get a list of selected database objects, we just need to list them after the "\d" command. I will list schemas, sequences and views. \dnsv

We can get lists of these objects: n – schemas f – functions s – sequences t – tables v – views     Just "\d" will give as tables and views by default. It is the same as "\dvt".

We can combine "\d" command with the letter "S". That will add system objects to lists. \dSt

"\dS" is equal to "\dSvt".

We can get SQL dump only for objects with the specific name. \dst tablo Name should be written with the lower letters, or inside of double quotes, like "Tablo".

We can use wildcards "*" (several letters) and "_" (one letter). \dst *o Wildcards can not be used inside of double quotes. This will not work: \dst "*o"

Query Execution Control

Query Formatting

We can control how the result of a query will be presented ( default is "sql" ). Let's run "SELECT * FROM sys.auths;" with different modes.

If we don't want to see the result of a query, we can use "trash" and "rowcount" modes.

\f trash --trash will return nothing

\f rowcount -- only row count will return

\f sql  --default

\f expanded  -- records are vertical

\f csv -- comma is separator

\f tab -- tab is separator

\f raw --raw database format

\f xml --like xml document

There are some special modes for the CSV.

\f csv=c --separator is letter "c", or any other sign

\f csv+c --header is included

<= Note that there are double quotes around the word "public".

For the rest =>, we will use query:   SELECT schema_path FROM sys.users;

\f csv --quotes are doubled

\f csv-noquote --no doubling

\f sql --default

The Number of Rows

When we run query that will return a lot or rows, we will not be able to see all of them. They will just whiz by before our eyes. We want to control how the rows are displayed. We want to see them page by page. For that we will use "\r" command.

For this example, I will use system table sys.function_types. It has 7 rows. SELECT * FROM sys.function_types;

I will set rows command to four: \r 4 We will get only one row for the first page. For the first page we always get N-3 rows. This is probably bug. If N is three or smaller, then the first page will be empty.   For other pages we will get the complete 4 pages. We can jump to the next page with "Enter". That is the same as "next".

If we type the letter "c" and then we press ENTER, we will jump to the end of a table.

If we type the letter "q" and then we press ENTER, we will immediately exit the execution.

For the last example I will make terminal window small.

I will set command to zero: \r 0 I will query this table. SELECT * FROM sys.keywords;

This is the best mode, because now the page is the size of the window.

I will press ENTER several times to go through a few pages. Then I will enlarge the window.

Only the first page will have 2 rows that are visible in one window. Other pages will have five rows. Obviously we have some "offset by 3" bug.

We will exit paging mode by setting the command to minus one "\r -1".

Width of a Table

If we set "\w" to minus one, MonetDB will think that we have unlimited width terminal. I've narrowed the window, so presentation will be squeezed and crumbly. It becomes better if window is wider.

Default value of the "\w" command is zero. I made my terminal window narrow, so in this mode 11/16 columns will not be shown. Only the width of the terminal is used.

If we set "\w" to some positive number, MonetDB will think that the terminal is only a certain number of characters wide. I'll set it to 50 characters, which is enough for 4 columns.

Auto Commit Mode

By default, auto commit mode is turned ON in mclient. We can turn it OFF with the command "\a". We turn it back to ON with "\A".

\a    INSERT INTO tablo( Number ) VALUES ( 1 ); ROLLBACK; SELECT * FROM tablo; --nothing was auto commited

\A INSERT INTO tablo( Number ) VALUES ( 2 ); ROLLBACK;      --not allowed in auto commit mode SELECT * FROM tablo;    --value 2 is auto commited

Logging and Monitoring

I will leave the session and then I will log in, just to clear the history. quit mclient -u monetdb -d DBfarmDB --password monetdb

I will turn on logging and tracing on mclient. This is mostly useful for debugging. \L /home/fffovde/Desktop/FileLog   --log to this file \X    --trace everything that mclient is executing

I will now execute a query: SELECT 2;

This is tracing. We'll get a lot of lines.

A new "LogFile" will appear with the similar content. This is logging.

We can disable logging and tracing. \L --we use \L without giving a full path \X false

I will run a few more queries: SELECT 3; SELECT 4;

Using the "\history" command, we can list all the SQL and commands we run in this session.

I will again quit the session, and log again, so I can change current directory.

quit cd /home/fffovde/Desktop mclient -u monetdb -d DBfarmDB --password monetdb

SQL Input and Output

We can redirect the result of our statement to some shell command. \| grep "alpha" SELECT * FROM (VALUES('alpha'),('beta'));

I will disable this command like this. Now our result is normal. \|     --without shell command SELECT * FROM (VALUES('alpha'),('beta'));

We can redirect the result of our statement to a file. \>ResultFile SELECT 10, 20;

Every statement we run will now go to the file, and not to the terminal. The file is in the current directory. We can disable the effect of  this command with: \>       --without the name of a file

On the Desktop, I will create one textual file with one query inside of it.

I can call this query from the file. I'll type: \<SqlFile

This is usefull when we want to run some prepared SQL script. The problem is that we will get the result, but we will not know what statement caused that result. To solve this we will use echo "\e".

After we turn on echo, every statement will be shown before the result of the statement. Useless for SQL but great for files.

SELECT 107;

\<SqlFile

I will again quit, and then login back, to disable effect of the echo command. quit mclient -u monetdb -d DBfarmDB --password monetdb

Other Backslash Commands

"\?" will give a list of all of the backslash commands.

"\help" will give a list of SQL statements.

We already saw how to use timing in this blog post "link". It is also possible to turn on timing inside of the mclient. Options are "none, clock, performance". \t clock SELECT 599;

\t performance SELECT 599;

--statements-- sql = parsing opt = optimizing run = running clk = total time

\t will return the current mode. \t none To disable timing.

We can always exit mclient with the "\q", or "quit".

Dash Commands

When we log in to mclient, we can provide an IP address, port number, username and database name. There are many more of these arguments

# mclient -h 192.168.100.124 -p 50001 -u monetdb -d DBfarmDB

Already Known Commands

Many commands exist in both backslash and dash version. I will give you example of such commands here.
During this example I will alternately login and quit, because dash commands can be run only during the login.

This will create sql dump of the whole database. mclient -u monetdb -d DBfarmDB -D

This is the same as "\D" backslash command, that we saw earlier.

We can enable tracing and logging.   mclient -u monetdb -d DBfarmDB -X -L FileLog

The same as "\X" and "\L".

We will set that our queries return expanded format, we will turn on echo and the timing. mclient -u monetdb -d DBfarmDB -f expanded -t clock -e   Format and Time support all of the possible modes.

The same as "\f, \t, \e".

With "-a" switch we can turn off autocommit for the whole session: "mclient -u monetdb -d DBfarmDB -a".

When we run a query, it will be paged to 5 rows, and the visible width will be 50 characters.   mclient -u monetdb -d DBfarmDB -r 5 -w 50

We already saw this with commands "\r, \w".

We can apply shell command on all of our queries. Notice that pipe must be escaped.   mclient -u monetdb -d DBfarmDB -\| 'grep m'

The same as \| grep "m".

History, Null and Timezone

I will login again with some new options: mclient -u monetdb -d DBfarmDB -z -n "zero" -H

Option "-z" is for disabling the time zone. Statement: SELECT CURRENT_TIME; , will give me Greenwich time, not my local time.

Option "-n" will define how the null will be presented in the terminal. This is only valid when "-f" format is either "sql,csv, tab or expanded".

Default format is "sql", we have set nul to be "zero", so the null will be presented as a "zero". SELECT null;        --result is "zero"

In this session we have run two statements: --SELECT CURRENT_TIME; --SELECT null;

Because our session was started with the switch history "-H" both of these statements are recorded in the file: /home/fffovde/.mapiclient_history_sql

I will now quit this session, and then I will login again: quit mclient -u monetdb -d DBfarmDB -H

Because I have used "-H" switch, history of mclient statements will be available for me. By pressing keyboard arrows, I will list that history.

If I run one more statement, that statement will be added to "/.mapiclient_history_sql" file. So, when "-H" switch is active, new statements will be added to history file, and that history will be available to us. SELECT 55; We can delete the history by deleting directly from the ".mapiclient_history_sql" file.

Getting Data from URLs

I will create a new table and then I will try to get data from the internet to insert into that table. Mclient will refuse to do that. CREATE TABLE IndustryList( Industry VARCHAR(50) ); COPY INTO IndustryList FROM 'https://cdn.wsform.com/wp-content/uploads/2020/06/industry.csv' ON CLIENT;

I will quit the session, and then I will login with a remote "-R" switch. quit mclient -u monetdb -d DBfarmDB -R   --password monetdb COPY INTO IndustryList FROM 'https://cdn.wsform.com/wp-content/uploads/2020/06/industry.csv' ON CLIENT;

It will work now. I can read from "IndustryList" table.   SELECT * FROM IndustryList;

SQL Dump with Insert Into

Normal dump will create a script that will fill a table with the data from the Stdin. It will use "COPY INTO FROM stdin" statement. mclient -u monetdb -d DBfarmDB -D

If we want to get "INSERT INTO" statements, then we have to add switch "-N". mclient -u monetdb -d DBfarmDB -D -N

Running SQL Statements from the Shell

With "-s" switch we can run a statement, instead of logging in. mclient -u monetdb -d DBfarmDB -s "SELECT null;"

We can also run a script directly from the shell. We previously created this file. mclient -u monetdb -d DBfarmDB SqlFile -i

We can use them both: mclient -u monetdb -d DBfarmDB -s "SELECT null;" SqlFile

Many of the previously shown options will work with the shell query, too.   mclient -u monetdb -d DBfarmDB -s "SELECT null;" -n "zero"

I will add command "\f xml" in the SqlFile.

This will make our command to fail. mclient -u monetdb -d DBfarmDB SqlFile

We will add interactive switch "-i". It will work now.             mclient -u monetdb -d DBfarmDB SqlFile -i

DOTMONETDBFILE

We already talked about DOTMONETDBFILE environment variable, in the blog post about backup "link". All of that is valid for the mclient, too.

user=monetdb password=monetdb host=192.168.100.145 port=50000

While Monetdbd is a tool that works with DBfarm, Monetdb is a tool that works with the database. Monetdb can be used to create, lock, release, start, stop, and destroy databases. In reality, Monetdb will just delegate all the work to Monetdbd, and Monetdbd will be the one that will perform all the operations.

Before we learn something about Monetdb, we first have to start DBfarm. We have created DBfarm in the previous post, about monetdbd, "link". Inside of this farm we already have database DBfarmDB.

monetdbd start /home/fffovde/Desktop/DBfarm

Database Commands

I will create new database in DBfarm.

monetdb create dbA

Database will be created in the maintenance mode.

Database that is in the maintenance mode can only be started on localhost with this command: monetdb start dbA

I will login to this database and I will create a new user: mclient -u monetdb -d dbA    #password monetb

CREATE USER "userA" WITH PASSWORD '123' NAME 'Geriel Damdin';

I will quit the session, and then I will login as a newly created user.

quit   #session                          mclient -u userA -d dbA

While the database is in the maintenance mode, only admin can login to it.

I will now release the database. That will pull database out of the maintenance mode. monetdb release dbA mclient -u userA -d dbA  #user will now be able to login

Currently database is out of the maintenance mode. How can we place our database to maintenance mode again? For that we just have to lock the database before we stop it.

quit    #session      monetdb lock dbA   monetdb stop dbA

If database is locked before it is closed, then only the administrator will be able to work with it.

Sample Databases

For further examples, I will create three more databases. monetdb create dbB dbC dbD

I will then get them out of the maintenance mode. monetdb -q release dbB dbC dbD

Most of the commands can work on several databases at once. We just have to name all of the databases in a row.

Notice that after the release command we didn't get any confirmation message. That is because we have used "--quiet" option. This option will always suppress confirmation messages. It will not suppress error messages, for example if we run it again. =>

Database Status

I will show you how to get status of a databases. I will change the status of the dbA and dbB for this example.

monetdb start dbA monetdb lock dbB

monetdb status If we type the status command, we will get some information about DBfarm databases. Database dbA is "LR" – locked and running for 1 minute. This database never crashed, so its health is 100%. On average, database dbA was running for 56 minutes.

In the state column we can see that our database can be in one of these states. We can also see for how long is database in that state.

B – booting

R – running.

S – stopped.

C – crashed.

LS – locked and stopped.

LR – locked and running.

In the health column we can see ratio "( successful starts/stops ) / ( all starts/stops )". 100% means that we never had unsuccessful start/stop.

In the health column we can see the times (56 min and 24 min). This is the average time the database has been up and running.

monetdb status -c dbA dbC We can specify a database name (or several database names) to get the status for only them.

monetdb status -l dbA

If we use options "-c" and "-l" with the status command, we would get several more pieces of information for our databases.

monetdb status -s rl With the "-s" option we can limit the displayed databases to only those that are running and/or locked ("rl"). The options are b (booting), r (running), s (stopped), c (crashed), l (locked).

Monetdb Information

Option "-v" will return the version of a server. monetdb -v We can also use the long syntax. monetdb --version

This is how we can get help.  We can only use long syntax for help.   monetdb  --help

Snapshots

Snapshot to Stdout

This command would write snapshot of a database to stdout. monetdb release dbA                         #first we have to release dbA monetdb snapshot write dbA	Our stdout is terminal, so we can not send binary data to stdout.
We can redirect stdout to a file. That way we can use "write" to get snapshot. monetdb snapshot write dbA > /home/fffovde/Desktop/dbA.tar

Snapshot to File

We can not create a snapshot in a file if property "snapshotdir" is not set. We do it with monetdbd. monetdbd set snapshotdir=/home/fffovde/Desktop /home/fffovde/Desktop/DBfarm

For "snapshot create" we have to provide full path, and that location must be inside of the "snapshotdir". I will try to backup database dbA. monetdb snapshot create -t /home/fffovde/dbAsnap dbA

I will try with this full path, that is inside of the "snapshotdir". This will work. monetdb snapshot create -t /home/fffovde/Desktop/ZZZ/dbAsnap dbA

With "-t" option we can create a snapshot file of one database, with the custom name. Folder "ZZZ" must be made in advance.

Without "-t" option, snapshot will be created with default settings. monetdb snapshot create dbB		We can not snapshot "tbB" because it is locked.
After we release the file, we will create the snapshot. monetdb release dbB monetdb snapshot create dbB	dbB_20250827T180139UTC.tar.lz4	Default name is made of the database name and the timestamp.
We can use wildcards for the name of a database. In this way we can create snapshots of several databases at once.	monetdb snapshot create db*

Restore a Snapshot

monetdb snapshot list		This is how we can get a list of all of the possible snapshots. For database dbB, we have two snapshots. Only default named snapshots are listed. Snapshot "dbAsnap" is missing because it has custom name.
We can restore second snapshot of the "dbB".	monetdb snapshot restore dbB@2

We can restore database under some other name. monetdb snapshot restore dbB@2 dbBBB

Inside of the "DBfarm" we now have this database twice.

We can also reference snapshot by its full path. monetdb snapshot restore -f /home/fffovde/Desktop/dbD_20250827T180619UTC.tar.lz4 dbD	In this case, we have to provide under which name we are restoring.
	Thanks to "-f" switch, we don't have to confirm overwrite of the current file.

Delete a Snapshot

This is how we can delete some snapshot.   monetdb snapshot destroy dbD@1
We can delete all but the last N snapshots for some database. monetdb snapshot destroy -f -r 1 dbB	Monetdb will delete the oldest dbB snapshot. We stil have one snapshot more.
	We can apply this command with the wild card, on all of the matched databases. I will leave zero snapshots, so we will delete them all. monetdb snapshot destroy -f -r 0 db*

We have no more snapshots that have default name. monetdb snapshot list

Database Settings

Read Database Settings

This is how we can read properties of some database. monetdb get all DBfarmDB   We can select what property we want to see: monetdb get optpipe DBfarmDB For each property, we will see its name, value, and whether that value is the default value.

name       prop              source      value DBfarmDB   name              -           DBfarmDB DBfarmDB   type              default     database DBfarmDB   shared            default     yes DBfarmDB   nthreads          default     10 DBfarmDB   ncopyintothreads  default     10 DBfarmDB   optpipe           default     default_pipe DBfarmDB   readonly          default     no DBfarmDB   embedr            default     no DBfarmDB   embedpy3          default     no DBfarmDB   embedc            default     no DBfarmDB   listenaddr        default     <unset> DBfarmDB   nclients          default     64 DBfarmDB   dbextra           default     <unset> DBfarmDB   memmaxsize        default     <unset> DBfarmDB   vmmaxsize         default     <unset> DBfarmDB   raw_strings       default     <unset> DBfarmDB   loadmodules       default     <unset>

Database Performance Settings

For most of the settings above, before we change them, we have to stop our database. Database "DBfarmDB" is currently stopped.

monetdb set nthreads=4 DBfarmDB monetdb set ncopyintothreads=6 DBfarmDB monetdb set optpipe=default_pipe DBfarmDB

Number of threads that database will use. By default, it is equal to number od CPU cores. Number of threads used for COPY INTO FROM. Default is equal to number of columns copied. We can find the list of possible optimizers in the system table "sys.optimizers".

After we change some property, the column "source" will be "local", instead of the "default". monetdb get nthreads DBfarmDB

I will now login to this DBfarmDB database. mclient -u monetdb -d DBfarmDB I will run this statement: select * from env() where name in ('gdk_mem_maxsize','gdk_vm_maxsize');

This is where we can see current values for properties "memmaxsize" and "vmmaxsize".

These values are in bytes, so maximal virtual memory size is about 4.000 terabytes, and maximal memory size is about 10.000 gigabytes. I will now exit and I will stop the database. quit monetdb stop DBfarmDB

#monetdb set memmaxsize=107374182400 DBfarmDB	This would limit RAM consumption of this database to 100GB. This is not hard limit; it is just suggestion to memory allocator.
#monetdb set vmmaxsize=107374182400 DBfarmDB	This is hard limit. Limiting virtual memory will limit how much RAM+Swap space this database can use.

Restriction Settings

monetdb set readonly=yes DBfarmDB monetdb set nclients=10 DBfarmDB monetdb set idletimeout=3600 DBfarmDB	We can set our database to be read only. We can limit number of the clients that can connect to our database. Connection will be closed if idle for this many seconds.
This property "idletimeout" is not listed when we use "get all" command, but inside of the file "DBfarmDB > .merovingian_properties" , that property is listed. It is also in documentation.

Extension Settings

#monetdb set embedpy3=true DBfarmDB #monetdb set embedr=true DBfarmDB #monetdb set embedc=true DBfarmDB

If we want to make UDF functions in python or "C" or "R" language, we have to set these settings to TRUE for specific databae. We used this for python in the blog post "link". For values we can use true/false or yes/no.

In this blog post "link", we have installed library "monetdb-odbc-loader". For such libraries to work in a database, we have to load them. This is done with "loadmodules" property.

#monetdb set loadmodules=odbc_loader DBfarmDB

Discovery Setting

I will now open green server. We have created that server in this blog post "link". monetdbd start /home/sima/monetdb/DBfarmG  #I will start daemon In the previous blog post "link", about monetdbd, we turned on discovery on the voc server for "DBfarm". I will again use discover command on the green server. monetdb discover

Discover command will show us all of the databases on the "voc" server, that are inside of the "DBfarm", and one local database on the green server.

We have turned on discovery for the whole "DBfarm" and for that we have used monetdbd. Now, I want to hide only one of those databases from discovery. I will hide database "DBfarmDB". monetdb set shared=no DBfarmDB

Discovery for "DBfarmDB" is turned off. Discover command will show us other databases from the "DBfarm", but not "DBfarmDB".

Other Settings

I will skip settings "type, dbextra, listenaddr". I don't understand them. I will talk about "raw_strings" setting.

I will login into DBfarmDB with the default password: mclient -u monetdb -d DBfarmDB

By default, the raw strings are turned OFF. I have talked about raw strings in this blog post "link". I will run some queries to see the default behaviour.

SELECT 'Bob\'s file';

SELECT 'a\nb';

SELECT 'C:\\temp\\new\\file.txt';

I will exit the session, and I will stop the database. quit monetdb stop DBfarmDB

I will change "raw_strings" property. monetdb set raw_strings=true DBfarmDB  #false is default Then, we will again login, and run those queries from above.

SELECT 'Bob''s file'; Backslash will not work.

SELECT 'a\nb';

SELECT 'C:\\temp\\new\\file.txt';

Off course, we can get the same results as before if we place prefix "E" in front of the strings. That will make them NOT raw.

I will quit session and stop the database.

quit monetdb stop DBfarmDB

Set Property to Default

Currently database can connect to the most 10 clients. We can change that back to the default value (64). monetdb get nclients DBfarmDB

monetdb inherit nclients DBfarmDB

For some properties we can unset them: monetdb set loadmodules=   DBfarmDB

Something More About Database Commands

We can provide password for monetdb user during database creation. monetdb create -p password123 dbE
This database will not be created in the maintenance mode. It has non-standard password, so there is no need for that. We can start it directly.	monetdb start dbE

Now, that our database is running, we can stop it with the "stop" command. This command will send SIGTERM signal, so it will kindly ask database to close, although the database can refuse that. Instead of the "stop" command, it is possible to us "kill" command. This command will send SIGKILL signal and the database will be violently killed.

I can kill "dbE" database by its name "monetdb kill dbE", or I can use "-a" switch to kill all of the open databases. monetdb kill -a

"-a" switch can be used with the commands "lock, release, start, stop and kill".

I will now delete this dbE database: monetdb destroy dbE

Remote Control

Remote control is explained in the blog post about monetdbd, "link". Take a look there to find out about switches "-h,-p,-P".

monetdb -h 192.168.100.124 -p 50000 -P zebra stop DBfarmDB

Monetdbd is a MonetDB management daemon. We use it to:
– Initialize a directory as a DBfarm. A DBfarm is a folder that can contain several MonetDB databases.
– Monetdbd is used to set and read farm-wide daemon options that are shared between databases.
– Monetdbd is used to start/stop the actual servers (Mserver5 process).
– Users usually connect to the monetdbd daemon first, and then this daemon will redirect them to the selected database.

Monetdbd Informations

monetdbd -v monetdbd --version

This is how we can find version of the monetdbd daemon. We can use short or long command syntax.

This is how we can get  help for monetdbd.   monetdbd -h monetdbd --help

We can provide the name of a command after the "-h" option. That will give us help for that command.

Monetdbd and DBfarm

cd /home/fffovde/Desktop mkdir DBfarm monetdbd create DBfarm

We will create a folder with the name "DBfarm". With the command "monetdbd create", we will initialize this folder. Inside of it, a new file ".merovingian_properties" will appear. Now it is possible to create new databases inside of this folder.

This is the inside of the ".merovingian_properties" file.

Currently no one is listening to default MonetDB port 50.000. I will make DBfarm directory functional.   monetdbd start DBfarm

Now our daemon is listening to port 50000, and DBfarm is functional. ss -tlnp | grep 50000

After the first start of the DBfarm, three new files will appear in the DBfarm directory. ".merovingian_lock" file is empty.

"merovingian.pid" file has process ID for monetdbd process. We can confirm that with "pgrep monetdbd".

"merovingian.log" file is a file with a log.

We can stop DBfarm with monetdbd.   monetdbd stop DBfarm

After closing the farm, file "merovingian.pid" will disappear. This file exists only when DBfarm is functional. Now, again, no one is listening to port 50.000. ss -tlnp | grep 50000

Getting and Setting DBfarm Settings

We can read setttings of the DBFarm.   monetdbd get all DBfarm

hostname              FffOvdeKomp dbfarm                DBfarm status                no monetdbd is serving this dbfarm mserver               unknown (monetdbd not running) logfile               DBfarm/merovingian.log pidfile               DBfarm/Merovingian.pid loglevel              information sockdir               /tmp listenaddr            localhost port                  50000 exittimeout           60 forward               proxy discovery             true discoveryttl          600 control               no passphrase            <unset> snapshotdir           <unset> snapshotcompression   .tar.lz4 keepalive             60 mapisock              /tmp/.s.monetdb.50000 controlsock           /tmp/.s.merovingian.50000

We can easily change any of these settings.  monetdbd set loglevel=debug DBfarm

We can also change the settings while DBfarm is active, but some changes will become valid only after Monetdbd is restarted.

After we first change some property, a list of the properties and their values will appear in the ".merovingian_properties" file. Not all of the properties that we can observer with the "get all" command will be listed here. Properties that do not appear here can not be changed with the "set" command. Those properties are informative and read only. We should never make changes in this file manually. We should always use the "set" command.

Monetdbd Settings

We will take a look at settings that we can set with Monetdbd.

 File and Log Settings

We know that in DBfarm folder we have files "merovingian.log" and "merovingian.pid". I will choose to use two other files instead of them.

I will change the settings for "log" and "pid" files: monetdbd set logfile=/home/fffovde/Desktop/logy.log DBfarm monetdbd set pidfile=/home/fffovde/Desktop/pidy.pid DBfarm I will start daemon again: monetdbd start DBfarm

Now, we have two new files on the desktop.

Inside of the "log" file, we have a log for monetdbd. Lines labeled with the MSG, are for information and warnings.

We can decide what will be logged in the log file. We can choose between modes "error, warning, information and debug".

error = only errors

warning = errors and warnings

information = errors, warnings and information

debug = the same as information

Error lines will be marked with ERR.

Default is "information". This is how we change the mode:

# monetdbd set loglevel=error DBfarm

Connection Settings

Settings "sockdir" and "forward" should almost never be changed.

Listendaddr is "localhost" by default. That means that our server is not available from the network.

If I set listenaddr to "0.0.0.0", our server will become available from the network. I will also change default port number to 50001.

monetdbd set listenaddr=0.0.0.0 DBfarm monetdbd set port=50001 DBfarm

This setting will become functional only after we restart monetdbd. monetdbd stop DBfarm monetdbd start DBfarm

Inside of the DBfarm, I will create and start one database: monetdb create DBfarmDB monetdb release DBfarmDB  monetdb start DBfarmDB

I will now open the green server. I explained creation of this server in this blog post "link". Then I will try to connect to DBfarmDB from the green server. mclient -h 192.168.100.124 -p 50001 -u monetdb -d DBfarmDB

We can use any of these modes for "listenaddr". It is also possible to use "hostname", in that case Linux will resolve what will be used.

	Loopback address:	External host:
only IPv4	127.0.0.1	0.0.0.0
only IPv6	::1	::
IPv4 and IPv6	localhost	all

I will change back the port number of the "voc" server to 50000: "monetdbd set port=50000 DBfarm".

Snapshot Settings

In the blog post "link", we have saw how to create a snapshot of our database. We have used stored procedure for that. The same is possible from the command line. Settings for command line snapshots are in monetdbd properties file.

"snapshotdir" is a directory where our snapshot will be made. If this is not set, then we will not be able to make a snapshot.	monetdbd set snapshotdir=/home/fffovde/Desktop DBfarm
"snapshotcompression" is a level of compression. We can choose between "tar.lz4,tar.bz2,tar.gz,tar.xz". This is optional setting.	monetdbd set snapshotcompression=.tar.lz4 DBfarm There is a dot before "tar".

I will set these changes, and then we will create a snapshot. We will create snapshot with the "monetdb" console program.

monetdb snapshot create DBfarmDB

This will create ".tar.lz4" snapshot on our Desktop. Inside of it we will get "DBfarmDB" database.

We can unset a property by providing an empty value:

monetdbd set snapshotdir= DBfarm

Remote Control Settings

We can call and control "monetdbd" on remote computer. I will allow remote control for "monetdbd" on the "voc" server.	monetdbd set control=true DBfarm
We also have to provide SHA512 password.	monetdbd set passphrase=zebra DBfarm
Our password will be written to ".merovingian_properties" file. It will be written as 128-digits hexadecimal number with the prefix "{SHA512}".
If we only have hashed password, then we can provide it directly.	monetdbd set passphrase={SHA512}9df34…24fec DBfarm
We have to restart monetdbd for these settings to become functional. We will also start DBfarmDB database.	monetdbd stop DBfarm monetdbd start DBfarm monetdb start DBfarmDB

Remote control is established by using monetdb.

I will run this command from the green server, with monetdb. Now we can control monetdbd on the "voc" server remotely.

monetdb -h 192.168.100.124 -p 50000 -P zebra stop DBfarmDB

Remote Discovery Settings

Monetdb can "shout" its existence on the local network by UDP broadcast. In regular intervals Monetdb server will tell other computers on the LAN about its existence. This is how we can discover all of the monetdb servers in our network. I will tell "voc" server to broadcast its existence every 300 seconds. It will use default port 50000. monetdbd set discovery=true DBfarm monetdbd set discoveryttl=300 DBfarm

I will restart DBfarm.   monetdbd stop DBfarm monetdbd start DBfarm

On the green server I will run this command. monetdb discover We can see here that on the "voc" server (hostname fffOvdeKomp), we have database "DBfarmDB". We also see local green database. So, we get a list of all of the available databases.

Start and Stop Settings

monetdbd stop DBfarm monetdbd start -n DBfarm

I will stop "DBfarm", and then I will start it with a "-n" switch.  This will leave "monetdbd" daemon in the foreground and will not return control to the shell (we will not get prompt in the next line). This is useful for debugging. We can close "monetdbd" using "ctrl+C" and then Enter.

There is "monetdbd" setting "exittimeout" with the default value 60.

This means that "monetdbd" will ask "mserver" to shutdown politely with the SIGTERM command ( "mserver" can refuse ). If the "mserver" is not closed after 60+5 seconds, then "monetdbd" will send SIGKILL signal, and "mserver" will be violently shut down.

There are three possible values for "exittimeout".

POSITIVE VALUE: "Mserver", if alive, will be killed 5 seconds after the "exittimeout".

NEGATIVE VALUE: "Mserver" will only receive SIGTERM signal. It will never be killed.

ZERO: "Monetdbd" will close, but "Mservers" will not. We should not use this.

When we wait for a really long query to finish, our client will not communicate with the server. Firewall can interpret that like an inactive connection, so it will close the connection. We want to avoid this, so we can use probe signal to keep connection alive.

This is done with the "keepalive" setting, with the default value of 60 seconds. # monetdbd set keepalive=60 DBfarm This probe signal will be sent exactly 127 times. If "keepalive" time is 60 seconds, that means that are connection will be at least (127 x 60 seconds) ≈ 2 hours alive.

Monetdbd Return Value

Monetdbd will return "0" for successful action, and "1" for unsuccessful. If we try to start already started daemon, we will get an error. monetdbd start DBfarm; echo $?

Monetdbd Process Signals

I will start the monetdbd again, and then I will read process ID of our daemon. monetdbd start DBfarm cat ./pidy.pid

Let's assume that I want to use some other file, instead of "logy.log", for logging, but I do not want to restart monetdbd. I will do that in the two steps.

monetdbd set logfile=/home/fffovde/Desktop/DBfarm/merovingian.log DBfarm kill -HUP  "$(cat ./pidy.pid)"

I am changing the file for my log. I am sending SIGHUP signal to monetdbd.

Now the daemon will stop using "logy.log" file, and will use "merovingian.log" file. We didn't have to restart the daemon.

We can kill monetdbd with a signal. All of the signals bellow work the same. kill -TERM "$(cat "pidy.pid")" kill -INT  "$(cat "pidy.pid")" kill -QUIT "$(cat "pidy.pid")"

They will ask our "mserver5" process to gently end (it can refuse). Signals: SIGTERM SIGINT SIGQUIT

File Backup

Our database ( "voc" ) is just a folder inside of the DBfarm folder. We can backup our database by making a copy of the "voc" folder. The problem is that before doing this we have to wait for the users to finish their queries, and then to prevent them from running new queries.

Disconnecting Users

We can check other users by reading form the sessions table. – Column "idle" is timestamp for session "1". This is the moment when that session finished running its last statement. – Column "idle" is null for session "0". In this session we are running current "SELECT … FROM sessions" statement, so this session is busy.

select sessionid, username, idle, login from sys.sessions;

If we estimate that database is not overloaded, then we can assume that all of the existing queries will end in 30 seconds. We will let existing transaction to run 30 seconds more, and then we will close all the connections.   call sys.shutdown(30); This procedure can be called only by the administrator.

If the "voc" user try to run a query, he will get "unexpected end of file" and his session will end. He will not be allowed to login again.

Shutdown procedure is using SIGTERM command. This procedure will ask database to shutdown, but the database can refuse to do it. In such circumstances we can use the second argument "call sys.shutdown(30, true);". Now we use SIGKILL command, and database will not be able to refuse shutting down.

After calling "sys.shutdown()", admin will still be able to run statements, not only during 30 seconds, but until he decides to leave the session. If admin quit the session, he will not be able to log into the server again.

At this moment, our database is not locked, but no one can connect to it.

monetdb status -l voc

Connection to this database will be possible only after the restart. But, after stopping the database, we want to prevent users from starting it again. We will lock our database so that only admin can start it again.   monetdb lock voc

The last step is to fully stop our database:   monetdb stop voc

Restoring Backuped File

We now just have to create a copy of the database folder. This copy can be placed on the same computer or we can move it to some other computer.

We can now release the new database and we can start it. monetdb release vocCopy mclient -u monetdb -d vocCopy

Cleaning Up After File Backup

We will quit and stop database vocCopy. quit monetdb stop vocCopy Now we can delete this vocCopy database: monetdb destroy vocCopy

Original "voc" database is still in the maintenance mode. We will release it with "monetdb release voc".

Hot Backup of a Database

A hot backup is the same as a file backup, but we don't have to stop the database. We'll make a copy of the database files while our database is still running.

Local Copy

As admin, we will call "hot_snapshot" procedure. The result will be our "voc" folder packed inside of the "tar" format. mclient -u monetdb -d voc    #password monetdb call sys.hot_snapshot('/home/fffovde/Desktop/backup.tar');
It is also possible to create compressed tar file. We can use ".lz4, .gz, .bz2, .xz" compression.   call sys.hot_snapshot('/home/fffovde/Desktop/backup.tar.lz4');

We can grant privileges of using our procedures to any user.

grant execute on procedure sys.hot_snapshot(string) to voc; grant execute on procedure sys.shutdown(tinyint) to voc;

Remote Copy

We will now backup database from the green server. Green server allows external access from the network. I explained how to create green server in this blog post "link".

We will open the green Database, on the green server. monetdbd start /home/sima/monetdb/DBfarmG monetdb start DatabaseG

From the "voc" server, we will login to the green server: mclient -h 192.168.100.145 -p 50000 -u monetdb -d DatabaseG

We will call "hot_snapshot", but we will provide second argument "false". That means that backup shouldn't be created on the server, but on the client. Backup of the green database will appear on the "voc" server. call sys.hot_snapshot('/home/fffovde/Desktop/backupGreen.tar', false);

Restoring Hot Backup

First, we will extract "voc" folder from the "tar" or compressed file.

I will place folder "DatabaseG" into farm folder on the "voc" server, and I will log into it. mclient -u monetdb -d DatabaseG  #password monetdb Now, we have green database on the "voc" server.

Cleaning Up After Hot Backup

I will quit green database session, and I will delete this database on the "voc" server. quit monetdb stop DatabaseG monetdb destroy DatabaseG