---
gitea: none
include_toc: true
---
# MySQL

Für **Joins**, siehe auch: [Roger SQL Join Tutorial](files/Roger_SQL_Join_Tut.docx)

---

## System

### Which MySQL is running?

Für 5.7.x versionen... Percona, Oracle...

``` sql
SELECT SUBSTR(variable_value,1,
LOCATE(' ',variable_value) - 1) DBVersion
FROM information_schema.global_variables
WHERE variable_name='version_comment';
```

### Show global variables

``` sql
SHOW GLOBAL VARIABLES;
SHOW GLOBAL VARIABLES LIKE 'innodb%';
SHOW GLOBAL VARIABLES LIKE 'innodb_rollback_on_timeout';
```

### Set global variables

Only until next MySQL restart. For permanent change, use my.cnf under the [mysqld] header. 

``` sql
SET GLOBAL innodb_lock_wait_timeout = 320;
```

---

## Users, Grants

### set root password

``` sql
ALTER USER 'root'@'localhost' IDENTIFIED WITH caching_sha2_password BY 'password';
FLUSH PRIVILEGES;
```

siehe auch https://linuxhint.com/change-mysql-root-password-ubuntu/

### Show users

``` sql
SELECT host,user,authentication_string FROM mysql.user;
```

### Add user

``` sql
CREATE USER 'newuser'@'localhost' IDENTIFIED BY 'password';
```

version 8.0

``` sq
nCREATE USER 'newuser'@'localhost' IDENTIFIED WITH mysql_native_password BY 'password';
```

### Show grants for user

``` sql
SHOW GRANTS FOR 'user_name'@'host';
```

### Grant

also see: https://www.digitalocean.com/community/tutorials/how-to-create-a-new-user-and-grant-permissions-in-mysql#how-to-grant-different-user-permissions

``` sql
GRANT ALL PRIVILEGES ON *.* TO 'newuser'@'localhost';  # give full root privileges
GRANT ALL PRIVILEGES ON db1.* TO 'newuser'@'localhost';  # give privileges for database db1
FLUSH PRIVILEGES;
```

### Revoke

``` sql
REVOKE type_of_permission ON database_name.* FROM 'username'@'localhost';
FLUSH PRIVILEGES;
```

### Change Password for user

``` sql
ALTER USER 'user-name'@'localhost' IDENTIFIED BY 'NEW_USER_PASSWORD';
FLUSH PRIVILEGES;

# or

UPDATE mysql.user SET authentication_string = PASSWORD('MyNewPass')
WHERE User = 'username' AND Host = 'localhost';
FLUSH PRIVILEGES;
```

---

## Database

### Drop DB

``` sql
DROP DATABASE IF EXISTS database_name;
```

### Export DB

```
mysqldump -h mariadb-p01.infra.vs.ch -P 33003 -uingest -p --routines --lock-tables=false --single-transaction atom > ~/atomp-01-20210425.sql
```

Windows

```
mysqldump -h mariadb-p01.infra.vs.ch -P 33003 -uingest -p --routines --lock-tables=false --single-transaction atom --result-file="c:\atomp-01-20210425.sql"
```

Falls Fehler `mysqldump: Couldn't execute 'SELECT COLUMN_NAME, JSON_EXTRACT.....` diese Option noch mit einbeziehen 
`--column-statistics=0`

See also https://www.linode.com/docs/guides/use-mysqldump-to-back-up-mysql-or-mariadb/

### Import DB

```
mysql -uusername -p db_name < /path/to/file.sql
```

Falls komprimierter dump:

```
gunzip < file.sql.gz | mysql -uusername -p db_name
```

### Create DB

```
mysql -h localhost -u root -p -e "CREATE DATABASE atom CHARACTER SET utf8 COLLATE utf8_unicode_ci;"
```

### Get Database Properties

``` sql
SELECT default_character_set_name, default_collation_name FROM information_schema.schemata WHERE schema_name = 'dar416';
```

---

## Tables

### Count rows in all tables

``` sql
select sum(TABLE_ROWS) from information_schema.tables where table_schema = 'db-name';
```

### Count rows in each table

``` sql
SELECT table_name, table_rows FROM INFORMATION_SCHEMA.TABLES WHERE TABLE_SCHEMA = 'db-name';
```

### Export/dump single table

```
mysqldump db_name table_name > table_name.sql
```

### Import single table

```
mysql -u username -p db_name < /path/to/table_name.sql
```

### Drop Tables

``` sql
DROP TABLE IF EXIST [table 1], [table 2], [table n];
```

### Select over single table

#### SELECT mit GROUP BY und HAVING

``` sql
SELECT post_title, COUNT(*) c
FROM wp_posts
WHERE post_type='article'
  AND post_author=1
GROUP BY post_title
HAVING c > 1
```

#### Select to file

````
mysql -uroot -e "use atom_A; select io.identifier, io.id, information_object_i18n.title from information_object io left join information_object_i18n on io.id = information_object_i18n.id where io.identifier IS NULL or io.identifier like '#%' or io.identifier = '?' order by identifier;" > /home/ubuntu/identifier_not_number.csv
```

#### List rows that have the same value in a column

``` sql
SELECT ve.vrzng_enht_id, vebhltn.bhltn_id
FROM vws_vrzng_enht_grund_daten  ve,
     vws_vrzng_enht_bhltn        vebhltn
WHERE vebhltn.vrzng_enht_id   =  ve.vrzng_enht_id
AND vebhltn.bhltn_id IN(
  SELECT vebhltn.bhltn_id FROM vws_vrzng_enht_bhltn vebhltn
  GROUP BY vebhltn.bhltn_id HAVING COUNT(*) > 1
)
```

### List duplicates (grouped), with additional col "found duplicates"

``` sql
SELECT post_title, COUNT(post_title)
FROM wp_posts
WHERE post_type = 'article'
  AND post_status = 'publish'
  AND post_title NOT LIKE 'Der Chart des Tages'
GROUP BY post_title
HAVING COUNT(post_title) >= 2;

# or, if all fields should be visible:
SELECT *, count(post_title)
FROM wp_posts
WHERE post_type = 'article'
  AND post_status = 'publish'
  AND post_title NOT LIKE 'Der Chart des Tages'
GROUP BY post_title
HAVING count(post_title) >= 2
```

### Select over multiple tables

#### SELECT with LEFT JOIN

``` sql
SELECT fuw_master.*, fuw_master_variation.name
FROM fuw_master
  LEFT JOIN fuw_master_variation
  ON fuw_master.id = fuw_master_variation.id_master;
```

### Create

#### Table with columns from 2 tables (using JOIN)

``` sql
CREATE TABLE fuw_master_2
SELECT fuw_master.*, fuw_master_variation.name
FROM fuw_master
  LEFT JOIN fuw_master_variation
  ON fuw_master.id = fuw_master_variation.id_master;
```

### Insert

``` sql
INSERT INTO blabla (spaltenname1, spaltenname2) VALUES ('wert1', 123)
```

### Update

``` sql
UPDATE table_name SET field1 = new-value1, field2 = new-value2
[WHERE Clause]
```

With replace:

``` sql
UPDATE table_name SET field_name = REPLACE(field_name,
        string_to_find,
        string_to_replace)
[WHERE Clause]
```

With joins:

``` sql
UPDATE information_object_i18n AS io 
  LEFT JOIN property p1 
    ON p1.object_id = io.id 
  LEFT JOIN property_i18n p2 
    ON p1.id = p2.id 
  SET io.title = CONCAT('recovery_failed_', io.title) 
WHERE p1.name LIKE 'PID' 
AND p2.value = 'CH-000-0:1234';
```

### Export from command line

```
mysql>SELECT * FROM tablexy 
INTO OUTFILE 'path/outfile.csv'
FIELDS TERMINATED BY ','
ENCLOSED BY '"',
ESCAPED BY '',
LINES TERMINATED BY '\n'
```

`path` muss meistens dem Pfad entsprechen, der unter `select @@secure_file_priv;` gesetzt ist!

---

## Version Upgrade 5.7 to 8.0 ##

  - https://dev.mysql.com/doc/refman/8.0/en/upgrade-binary-package.html
  - https://tastethelinux.com/2020/09/14/upgrade-mysql-server-from-5-7-to-8-ubuntu-18-04/

## Install 5.7 on Ubuntu 20 

  - https://www.vultr.com/docs/how-to-install-mysql-5-7-on-ubuntu-20-04/

## Fehlermeldungen 

### Error "1038 Out of memory, consider increasing server sort buffer size

  - https://www.skynats.com/blog/error-1038-mysql-memory-allocation-error/

### Disk voll wegen zu vielen binlogs

In `/var/lib/mysql` sind zu viele binlogs. Lösung: auf nur 1 Woche reduzieren. Binlogs werden für Restorte gebraucht.

  - mysql service stoppen
  - `/etc/mysql/conf.d/mysql.cnf` erweitern um: `binlog_expire_logs_seconds=604800`
  - mysql neu starten