Documentos de Académico
Documentos de Profesional
Documentos de Cultura
Silo - Tips Informix Ultimate Warehouse Edition
Silo - Tips Informix Ultimate Warehouse Edition
Agenda
Agenda
Base de Datos para Transacciones (OLTP) Data Warehouse para Análisis (OLAP)
Soporta las operaciones de misión crítica y día-a-día del negocio Soporta la estrategia y toma de decisiones del negocio
Acceso a través de aplicaciones específicas Acceso a través de herramientas de inteligencia de negocios
Diseño de BD a partir de la aplicación (Modelo Entidad-Relación) Diseño multidimensional: métricas y perspectivas del negocio (Estrella)
Data actual, aislada a un sistema, detallada, normalizada Data histórica, integrada, detallada/sumarizada a cierto nivel, duplicada
Muchos usuarios concurrentes (de niveles operativo e IT) Pocos usuarios concurrentes (de niveles gerencial y ejecutivo)
Transacciones cortas y simples, lecturas y cambios aleatorios en línea Queries complejos, joins, filtros, agregaciones; Delta de cargas (batch)
Acceso aleatorio a poco volumen de datos, operaciones sobre registros Acceso secuencial a gran volumen de datos, operaciones sobre columnas
Uso repetitivo, SQL pre-definidos y pre-optimizados Uso ad-hoc, queries no conocidos ni optimizados a priori
Métrica de rendimiento: Transaction Throughput Métrica de rendimiento: Query Throughput & Respuesta
4 © 2011 IBM Corporation
Software Group – Information Management
Retos:
– Lentitud en la carga completa e incremental del warehouse: Cada registro a insertar es separado en sus columnas,
cada una se almacena en forma comprimida junto con las otras instancias de valores de dicha columna
– Lentitud en operaciones basadas en registro u obtención de múltiples columnas: Cada vez que se requiere un registro
se obtienen y descomprimen separadamente los valores de sus columnas, y se agrupan para formar el registro
– Tienen necesidad de producir mecanismos internos para “imitar” un sistema relacional en varios aspectos
– Puede necesitarse: disponer de mucha memoria para escalar, crear índices apropiados, y reorganizar toda la data
– No apropiados para ambientes mixtos OLTP/OLAP
– Los datos pueden no estar disponibles durante las cargas/actualizaciones
– Dependiendo de la implementación, pueden ofrecer limitaciones adicionales en: el esquema y jerarquía de datos,
acceso a los datos, uso de estándarles y SQL, alta-disponibilidad, escalabilidad, concurrencia de usuarios, etc
Key Findings:
Data warehouse DBMSs have evolved to a broader analytics infrastructure supporting
operational analytics, corporate performance management and other new applications and
uses.
Cost is driving interest in alternative architectures but performance optimization is driving
multi-tiered data architectures and a variety of deployment options - notably a strong interest
in in-memory data mart deployments.
Analysis:
Optimization techniques such as summaries, aggregates and indexes are simply the result
of performance restrictions inherent to normalized data and the way the RDBMS manages
rows and columns.
Column-Store DBMSs:
CIOs should be aware that their current DBMS vendor may offer a column-store solution. Don’t just buy a
column-store-only DBMS because a column store was recommended by your team.
In-Memory DBMSs:
IMDBMS technology also introduces a higher probability that analytics and transactional systems can
share the same database.
Agenda
Informix
Warehouse
Accelerator
Cognos integration
- Native Content Store on IDS
SQL Merge
Informix Warehouse with
Informix Warehouse Storage
Feature Optimization/Compression
- SQW
- Data Modeling
- ELT/ETL
Agenda
(DevWorks Article) Compare the Informix Version 11 editions (by Carlton Doe)
Learn the differences in features, functions, and pricing models
http://www.ibm.com/developerworks/data/library/techarticle/dm-0801doe/index.html
(Brochure) Informix product family V11.7 (1.39MB)
ftp://public.dhe.ibm.com/common/ssi/ecm/en/imb14078usen/IMB14078USEN.PDF
(Overview Page) IBM Informix Product family
http://www.ibm.com/software/data/informix/
+
IBM Informix Warehouse Accelerator
(Intel Linux 64-bit hardware + software)
Agenda
5 3
Predicate evaluation on 4 Frequency Partitioning
compressed data Enabler for the effective parallel access of
Often scans w/o decompression the compressed data for scanning.
during evaluation Horizontal and Vertical Partition
Elimination.
Massive Parallelism
All cores are used within cpus for queries
A MART is a logical collection of tables which are related to each other. For example, all tables of a single
star schema would belong to the same MART.
The administrator uses a rich client interface to define the tables which belong to a MART together with
the information about their relationships.
IDS creates definitions for these MARTs in the own catalog. The related data is read from the IDS tables
and transferred to IWA.
The IWA transforms the data into a highly compressed, scan optimized format which is kept locally (in
memory) on the Accelerator
IDS + IWA
Coordinator Worker
Process Processes
Define
30
30 accelerator storage directory: compressed©data
2011 IBM Corporation
on disk
Software Group – Information Management
TCP/IP
IDS SQL Query
Processor
Query Router Results
(via DRDA)
Compressed
DB partition
The data marts or subset of data marts loaded into the IWA are kept by Informix as Accelerated Query
Tables (AQT)
– AQT = A view definition that will allow the Informix Optimizer to recognize which queries could be
accelerated via IWA
Turning Acceleration ON:
– SQL statement “set environment use_dwa ‘1’”
Turning Acceleration ON with debugging in Online.log:
– SQL statement “set environment use_dwa ‘3’”
Turn Acceleration OFF:
– SQL statement “set environment use_dwa ‘0’” (DEFAULT)
The SQL explain output : New information
– (IWA-Executed) if the query was successfully matched against an AQT and therefore offloaded to the
accelerator
– (IDS-Executed) if the query did not match an AQT or could not be accelerated through IWA and was
processed by the Informix Dynamic Server (IDS) alone
– (IWA-Avoid_Execute) if the query was invoked under “SET EXPLAIN ON AVOID_EXECUTE” for the
plan to be generated by the Informix Optimizer without actually executing the query; and if it was
executed, it would be offloaded to IWA. If (IDS-Avoid_Execute) is produced instead, the query was to
be resolved by IDS only.
SQL Tracing will generate additional information regarding query acceleration in the “onstat
-g his” output. Its corresponding catalog table sysmaster:syssqltrace has these new
columns:
– sql_aqtinfo: a value > 0 is the time (in ms) spent in query matching; a value of 0 means
the SQL statement did not qualify for offloading;
– and a value < 0 is the acceleration error code obtained
– sql_aqt: contains the name of the AQT matching the query
Opciones:
Agenda
Our Retail users will be really happy to see such a huge improvement in the
queries processing timings.
This IWA extension to IDS will really bring value to the Retail BI environment.
Agenda
Videos about Informix Warehouse Accelerator (based on Fred Ho's webcast / presentation):
Informix Warehouse Accelerator
http://www.youtube.com/watch?v=ohyanfPucE4
http://www.youtube.com/watch?v=ohyanfPucE4&feature=related
Part 1: Informix warehouse Overview.
http://www.youtube.com/watch?v=MILUTXHez-M&feature=related
Part 2: Informix Warehouse - SQW tool
http://www.youtube.com/watch?v=R_BajAaxR2A&feature=related
Part 3: Informix v11.70 warehouse features overview.
http://www.youtube.com/watch?v=Tqtn3Rq6G0U&feature=related
Part 4: Informix warehouse accelerator. Overview and integration
http://www.youtube.com/watch?v=O6DzOfd7uC4&feature=related
Part 5/5: Informix Warehouse Accelerator technology.
http://www.youtube.com/watch?v=7PXRKIkKkRU&feature=related
Links (at PlanetIDS site) to Set up tips for Informix Warehouse Accelerator:
http://planetids.com/
– Informix Warehouse Accelerator - Using the Eclipse based Administration Tool
– Informix Warehouse Accelerator - Configuration Tips - Part 5 - Size of /dev/shm
– Informix Warehouse Accelerator - Configuration Tips - Part 4 - Understanding Linux Out of Memory
– Informix Warehouse Accelerator - Configuration Tips - Part 3 - Linux Kernel Tuning
– Informix Warehouse Accelerator - Configuration Tips - Part 2 - Memory Resources
– nformix Warehouse Accelerator - Configuration Tips - Part 1 - CPU Resources
Blogs:
– Fred Ho's blog at IBM DeveloperWorks
https://www.ibm.com/developerworks/mydeveloperworks/blogs/fredho66/?lang=en_us
– Keshav's blog at IBM DeveloperWorks
https://www.ibm.com/developerworks/mydeveloperworks/blogs/Keshav/?lang=en
Preguntas?