Monitoreo de MySQL y PostgreSQL con SQL

POST GRESQL – M YSQL
INST RU CT OR: EM A NU EL CA LVO
Tips de monitoreo con SQL

C O NC EPT O S B Á SIC O S
Premisas

Premisas
 Evitar/monitorear la latencia de escritura y lectura.
 Es lo que tardan los dispositivos de almacenamiento, en
devolver los datos.
 Aumentar/monitorear el rendimiento de salida de
resultados.
 Capacidad de respuesta del servidor.
 Seguridad.
Bases de datos veloces,
implican aplicaciones
Veloces.

Como lograr …?
 Menor latencia de I/O:
 Más discos.
 Discos más rápidos.
 Tablespaces y particiones separadas entre si.
 Utilización de sistemas de ficheros más avanzados.
 Rendimiento:
 Más memoria.
 Más y mejores procesadores y núcleos.
 Mejor conexión entre servidores en red.
 Servidores dedicados.

Monitoreo Básico
 Desde el Sistema Operativo.
 Se obtiene el rendimiento en términos específicos del sistema
operativo y hardware.
 Desde el motor de Base de datos.
 Se puede obtener datos de accesos a objetos y cantidad de datos en
caché de los mismos.

Monitoreo de Base
 Estado de los accesos a relaciones.
 Estado de las estadísticas.
 Estados de los índices.
 Estado del caché.
 Procesos en ejecución.
 TPS con pgbench, medir rendimiento.
 Consultas lentas.

Postgresql Mysql
 Select
pg_size_pretty(pg_dat
abase_size(name));
 SELECT
pg_size_pretty(pg_tot
al_relation_size(tabla)
);
 SELECT table_schema "Data Base
Name", sum( data_length +
index_length ) / 1024 / 1024 "Data
Base Size in MB"
FROM information_schema.TABLES
GROUP BY table_schema ;
 SELECT table_schema "Data Base
Name",
sum( data_length + index_length ) /
1024 /
1024 "Data Base Size in MB",
sum( data_free )/ 1024 / 1024 "Free
Space in MB"
FROM
information_schema.TABLES
GROUP BY table_schema ;
Tamaño de la base y tablas

Postgresql Mysql
 SELECT
pg_size_pretty(pg_tot
al_relation_size(tabla)
);
 SELECT
pg_size_pretty(pg_rela
tion_size(tabla));
 SELECT table_name,
table_rows, data_length,
index_length,
round(((data_length +
index_length) / 1024 /
1024),2) "Size in MB"
FROM
information_schema.TA
BLES WHERE
table_schema =
"schema_name";
Tamaño tablas

Postgresql Mysql
 SELECT * FROM
pg_stat_activity;
 SHOW PROCESSLIST;
 SHOW STATUS LIKE
'%threads%';
 show session status like
'connections';
Procesos

A PA RT A D O D E C O NSU LT A S PA RA M YSQ L
Mysql

Engines (Mysql)
 Show engines;
 show engine innodb status;

InnoDB Status
Per second averages calculated from the last 59 seconds
-----------------BACKGROUND THREAD-----------------
srv_master_thread loops: 4 1_second, 0 sleeps, 0 10_second, 5 background, 5 flush
srv_master_thread log flush and writes: 0
----------SEMAPHORES----------
OS WAIT ARRAY INFO: reservation count 3, signal count 3
Mutex spin waits 1, rounds 30, OS waits 0
RW-shared spins 3, rounds 90, OS waits 3
RW-excl spins 0, rounds 0, OS waits 0
Spin rounds per wait: 30.00 mutex, 30.00 RW-shared, 0.00 RW-excl
------------TRANSACTIONS------------
Trx id counter 502
Purge done for trx's n:o < 32C undo n:o < 0
History list length 7
LIST OF TRANSACTIONS FOR EACH SESSION:
---TRANSACTION 0, not started, OS thread id 2988
MySQL thread id 6, query id 110 localhost 127.0.0.1 root
show engine innodb status
--------FILE I/O--------
I/O thread 0 state: wait Windows aio (insert buffer thread)
I/O thread 1 state: wait Windows aio (log thread)
I/O thread 2 state: wait Windows aio (read thread)
I/O thread 6 state: wait Windows aio (write thread)
Pending normal aio reads: 0 [0, 0, 0, 0] , aio writes: 0 [0, 0, 0, 0] ,
ibuf aio reads: 0, log i/o's: 0, sync i/o's: 0
Pending flushes (fsync) log: 0; buffer pool: 0
191 OS file reads, 7 OS file writes, 7 OS fsyncs
0.00 reads/s, 0 avg bytes/read, 0.00 writes/s, 0.00 fsyncs/s
--------------------INSERT BUFFER AND ADAPTIVE HASH INDEX-------------------
Ibuf: size 1, free list len 0, seg size 2, 0 merges
merged operations:
insert 0, delete mark 0, delete 0
discarded operations:
insert 0, delete mark 0, delete 0
Hash table size 195193, node heap has 0 buffer(s)
0.00 hash searches/s, 0.00 non-hash searches/s
---LOG---
Log sequence number 3308699
Log flushed up to 3308699
Last checkpoint at 3308699
0 pending log writes, 0 pending chkp writes
10 log i/o's done, 0.00 log i/o's/second
---------BUFFER POOL AND MEMORY----------------------
Total memory allocated 49971200; in additional pool allocated 0
Dictionary memory allocated 18671
Buffer pool size 3008
Free buffers 2829
Database pages 179
Old database pages 0
Modified db pages 0
Pending reads 0
Pending writes: LRU 0, flush list 0, single page 0
Pages made young 0, not young 0
0.00 youngs/s, 0.00 non-youngs/s
Pages read 179, created 0, written 1
0.00 reads/s, 0.00 creates/s, 0.00 writes/s
No buffer pool page gets since the last printout
Pages read ahead 0.00/s, evicted without access 0.00/s
LRU len: 179, unzip_LRU len: 0
I/O sum[0]:cur[0], unzip sum[0]:cur[0]
----------ROW OPERATION------------
0 queries inside InnoDB, 0 queries in queue
1 read views open inside InnoDB
Main thread id 796, state: waiting for server activity
Number of rows inserted 0, updated 0, deleted 0, read 11
0.00 inserts/s, 0.00 updates/s, 0.00 deletes/s, 0.00 reads/s

Explorando tablas
mysql> select * from
information_schema.tables
where TABLE_NAME like 'prueba' limit
1G
******* 1. row *****************
TABLE_CATALOG: def
TABLE_SCHEMA: mysql
TABLE_NAME: prueba
TABLE_TYPE: BASE TABLE
ENGINE: InnoDB
VERSION: 10
ROW_FORMAT: Compact
TABLE_ROWS: 5
AVG_ROW_LENGTH: 3276
DATA_LENGTH: 16384
MAX_DATA_LENGTH: 0
INDEX_LENGTH: 0
DATA_FREE: 0
AUTO_INCREMENT: 6
CREATE_TIME: 2010-11-
03 12:09:45
UPDATE_TIME: NULL
CHECK_TIME: NULL
TABLE_COLLATION:
latin1_swedish_ci
CHECKSUM: NULL
CREATE_OPTIONS:
TABLE_COMMENT:
1 row in set (0.00 sec)

Explarando Tablas (2)
 show table status from
mysql like 'prueba'G

Particiones
mysql> select * from
information_schame.partitions where
table_name = 'prueba' limit 1G
********** 1. row ***************************
TABLE_CATALOG: def
TABLE_SCHEMA: mysql
TABLE_NAME: prueba
PARTITION_NAME: NULL
SUBPARTITION_NAME: NULL
PARTITION_ORDINAL_POSITION: NULL
SUBPARTITION_ORDINAL_POSITION: NULL
PARTITION_METHOD: NULL
SUBPARTITION_METHOD: NULL
PARTITION_EXPRESSION: NULL
SUBPARTITION_EXPRESSION: NULL
PARTITION_DESCRIPTION: NULL
TABLE_ROWS: 5
AVG_ROW_LENGTH: 3276
DATA_LENGTH: 16384
MAX_DATA_LENGTH: NULL
INDEX_LENGTH: 0
DATA_FREE: 0
CREATE_TIME: 2010-11-03 12:09:45
UPDATE_TIME: NULL
CHECK_TIME: NULL
CHECKSUM: NULL
PARTITION_COMMENT:
NODEGROUP:
TABLESPACE_NAME: NULL

Estadísticas
mysql> select * from statistics
where table_name like 'prueba' limit 1G
********* 1. row ***************************
TABLE_CATALOG: def
TABLE_SCHEMA: mysql
TABLE_NAME: prueba
NON_UNIQUE: 0
INDEX_SCHEMA: mysql
INDEX_NAME: PRIMARY
SEQ_IN_INDEX: 1
COLUMN_NAME: the_key
COLLATION: A
CARDINALITY: 5
SUB_PART: NULL
PACKED: NULL
NULLABLE:
INDEX_TYPE: BTREE
COMMENT:
INDEX_COMMENT:

Working example
mysql> show table status like 'prueba'G
Name: prueba
Engine: InnoDB
Version: 10
Row_format: Compact
Rows: 9703
Avg_row_length: 35
Data_length: 344064
Max_data_length: 0
Index_length: 0
Data_free: 0
Auto_increment: 16371
Create_time: 2010-11-03 12:09:45
Update_time: NULL
Check_time: NULL
Collation: latin1_swedish_ci
Checksum: NULL
mysql> delete from prueba where a between
80 and 81; Query OK, 218 rows affected (0.30
sec)
Name: prueba
Engine: InnoDB
Version: 10
Row_format: Compact
Rows: 10823
Avg_row_length: 31
Data_length: 344064
Max_data_length: 0
Index_length: 0
Data_free: 0
Create_time: 2010-11-03 12:09:45
Update_time: NULL
Check_time: NULL
Checksum: NULL

Working example (2)
mysql> optimize table pruebaG
1. row ***************************
Table: mysql.prueba
Op: optimize
Msg_type: note
Msg_text: Table does not support
optimize, doing recreate + analyze
instead
2. row ***************************
Table: mysql.prueba
Op: optimize
Msg_type: status
Msg_text: OK
2 rows in set (2.13 sec)
1. row ***************************
Name: prueba
Engine: InnoDB
Version: 10
Row_format: Compact
Rows: 10508
Avg_row_length: 31
Data_length: 327680
Max_data_length: 0
Index_length: 0
Data_free: 0
Create_time: 2010-11-03 12:09:45
Update_time: NULL
Check_time: NULL
Checksum: NULL
Create_options:

A PA RT A D O C O NSU LT A S PA RA PO ST GRESQ L
Postgresql

Monitoreo de Base
 Estado de los accesos a relaciones.
 Pg_statio_user_tables
 Pg_stat_user_tables
 Tamaños
 select pg_size_pretty( pg_database_size(‘ejemplo'));
 select pg_size_pretty( pg_relation_size('datos'::regclass));

Monitoreo de Base
 Estado de las estadísticas.
 Pg_stats
 SELECT * FROM pg_stats WHERE tablename = ‘tabla’ AND
attname = ‘columna’;

Monitoreo de Base
 Estados de los índices.
 Pg_stat_user_indexes
 Pg_statio_user_indexes

Monitoreo de Base
 Estado del caché.
 Cantidad de bloques leídos:
 Select pg_stat_get_db_blocks_fetched((select datid from
pg_stat_database where datname = 'pampabs'));
 Bloques leídos y en caché:
 Select pg_stat_get_db_blocks_hit((select datid from
pg_stat_database where datname = 'pampabs'));

Postgresql Mysql
 Consultas lentas.
 Activar
log_min_duration_state
ment (milisegundos).
 Consultas lentas
 log_slow_queries
 log_queries_not_using_i
ndexes
Consultas lentas

Trucos
 Realizando un ALTER TABLE sin modificar la tabla
se obliga la reestructuración.
 Ordenar los registros físicamente:
 SELECT * INTO tabla2 FROM tabla ORDER BY columna;

A LGU NOS CONT RIB S PA RA M ONIT OREO
Contribs Postgresql

pg_stat_statements
 Loguea absolutamente
todas las consultas.
 Funciones:
 pg_stat_statements_reset
()
 Se utiliza:
 Select * from
pg_stat_statements;
#Configuracion en el Postgresql.conf:
shared_preload_libraries='pg_stat_state
ments'
custom_variable_classes='pg_stat_state
ments'
pg_stat_statements.max = 10000
pg_stat_statements.track = all
pg_stat_statements.save = on

pgfouine
 Descargar desde
pgfoundry.org.
 Necesita un php-cli.
 Si tenemos que usar
stderr, el log_line_prefix
debe establecerse en:
 ‘%t [%p]: [%l-1]’
Php pgfouine.php -format html-
with-graphics –logtype stderr –
file <archivo_log> >
result.html

pgstattuple
 Varias funciones de
medición estadística.
 Solo hacen un bloqueo de
lectura.
 pgstattuple(‘table’)
 pgstatindex(‘index’)
 pg_relpages

pgrowlocks
 Devuelve la información
por tupla.
 Devuelve lock_type.
SELECT *
FROM
general g JOIN
pgrowlocks(‘general’) p
ON (g.ctid = p.locked_row);

Gracias Por asistir
Esta presentación está
bajo licencia GPL.
Contacte a: postgres.arg
(at) gmail.com
Skype: emanuel.cfranco

Monitoreo de MySQL y PostgreSQL con SQL

Related slideshows

More Related Content

What's hot

What's hot (20)

Similar to Monitoreo de MySQL y PostgreSQL con SQL

Similar to Monitoreo de MySQL y PostgreSQL con SQL (20)

More from Emanuel Calvo

More from Emanuel Calvo (9)

Recently uploaded

Recently uploaded (20)

Monitoreo de MySQL y PostgreSQL con SQL