Technical white paper Innovative technology for big data analytics The HP Vertica Analytics Platform database provides price/performance, scalability, availability, and ease of administration Table of contents The time for innovation is now The HP Vertica Analytics Platform advantage Radically improved database price-performance Painless scalability DBA liberation Key innovations Faster query performance Store data in significantly less table space High performance and high availability High efficiency with recovery by query Load more data and deliver more realtime answers daily Query more data faster For more information
The time for innovation is now Regulatory compliance, increased competition, and other pressures mean you need to accumulate and analyze larger and larger quantities of data. Many companies now have hundreds of terabytes of data to store and analyze. Yet, most database management innovation has not kept pace. Performing ad hoc queries on such large data volumes does not come naturally for existing database management systems (DBMS), which use a row-oriented design for writeintensive transaction processing rather than for read-intensive analytics. Desperate for better performance, many roworiented DBMS customers spend millions of dollars every year on stop-gap measures, such as adding database administrator (DBA) resources, creating and maintaining OLAP cubes, or replacing their DBMS with expensive and proprietary data warehouse appliances. The HP Vertica Analytics Platform advantage The HP Vertica Analytics Platform, an affordable, limitlessly scalable analytics platform, includes a DBMS architecture that scales with your data. It provides real-time query performance for hundreds of gigabytes of data to hundreds of terabytes of data and keeps up when end-user requirements change rapidly. Let s take a closer look at what sets HP Vertica Analytics Platform apart. Radically improved database price-performance Benchmarks show that HP Vertica Analytics Platform can run 0-to-100 times faster than traditional DBMS and data warehouse appliances. HP Vertica Analytics Platform runs on commodity hardware. Couple that with its pricing approach, and you can deploy large-scale query-intensive databases for a fraction of the cost of traditional data management solutions.
Painless scalability HP Vertica Analytics Platform runs on standard, commodity hardware running Linux and is optimized for cost-efficient, distributed computing environments such as grids and clusters. HP Vertica Analytics Platform licensing is based on the amount of data you store not your database hardware configuration. This means you can scale the size and usage of your database without: Disruptive database platform changes Proprietary data warehouse appliances Paying incremental DBMS software licensing fees each time you add a CPU to the system DBA liberation We ve built a lot of DBA know-how into HP Vertica Analytics Platform to keep it running efficiently without much administrative overhead. High availability, disaster recovery, schema design, and physical optimization run automatically, freeing your DBAs to focus on higher value-added activities. Key innovations From a database developer s perspective, the database in the HP Vertica Analytics Platform looks pretty standard. It supports SQL, ACID transactions, and JDBC, and it works with popular ETL (extract, transform, and load) and business intelligence (BI) reporting products. Underneath the covers, it s a different story. The HP Vertica Analytics Platform database is designed to aggressively economize disk I/O and is written natively to support grid computing. It provides a modern solution for today s large-scale, read-intensive database applications, with ground-breaking architectural features. Faster query performance In a row-oriented DBMS, row values are stored contiguously. To process a query, the DBMS must read data from every row and column, even columns not specified in the query. Bitmap indices, data cubes, materialized views, and other database features can help. However, for many databases, managing data structures to improve performance for every query and use is prohibitively complex, and these data structures often impose dramatic storage space and update performance overhead. HP Vertica Analytics Platform stores the values for each column contiguously, so it only needs to read the columns being queried. This approach dramatically improves query performance by eliminating unnecessary disk and memory I/O. Store data in significantly less table space CPUs are getting faster at a much greater rate than disk bandwidth is increasing, so HP Vertica Analytics Platform replaces slower disk I/O with faster CPU cycles to encode data elements into a more compact form and query them. Its innovative query engine operates directly on compressed data, meaning that it can actually require fewer CPU operations to process the compressed version of a table. HP Vertica Analytics Platform supports logical relational models. It stores data physically as projections collections of sorted columns, similar to materialized views. Multiple projections stored on networked, shared-nothing machines, or sites, can contain overlapping subsets of columns with different sort orders for high availability and enhance performance by executing queries against the projections with the most appropriate columns and sort orders. Data physically stored as multiple projections: Collections of columns with multiple sort orders Distributed across multiple nodes Aggressively compressed Queries and loading parallelized across them Built-in redundancy for high availability Designed automatically
HP Vertica Analytics Platform compression methods High performance and high availability Based on DBA-provided logical schema definitions and SQL queries, HP Vertica Analytics Platform automatically determines what projections to construct and where to store them to optimize query database performance and high availability. High efficiency with recovery by query Rather than having a mirrored database backup sitting idle for failover purposes, HP Vertica Analytics Platform leverages the redundancy built into the database s projections. It queries projections not only for user requests, but also for rebuilding the data in a recently restored projection or site. The database designer builds necessary redundancy into the projections it creates so that a DBA-specified number of site failures can occur without compromising the system. This approach to recovery avoids bogging down database performance with expensive logging and two-phase commit operations. Load more data and deliver more real-time answers daily Many customers query their data management systems and warehouses by day and bulk-load them with data by night. The problem is, you have too much data to load at night and your users are demanding more real-time data. HP Vertica Analytics Platform features a hybrid architecture that supports querying and loading in parallel across multiple projections. Each HP Vertica Analytics Platform site contains a memory-resident write-optimized store (WOS) for recording inserts, updates, and deletes, and a read-optimized store (ROS) for handling queries. WOS contents continuously move into the associated ROS asynchronously. Lightweight transaction management prevents database reads and writes from conflicting, so you can run queries against data in the ROS, WOS, or both.
HP Vertica Analytics Platform hybrid architecture querying and loading never conflict. Query more data faster The improved database price/performance, scalability, availability, and ease of administration you get with HP Vertica Analytics Platform let you deliver more data to more people for more use at less cost and effort. HP Vertica Analytics Platform is ideal for all kinds and sizes of businesses, including applications such as: Data marts Business intelligence Data warehousing Click stream analysis Fraud detection Compliance reporting Call detail analytics Tick-store query applications Buying pattern analysis Basel II compliance Sales dashboards And more For more information To learn more about how the HP Vertica Analytics Platform can help your company perform data analytics more effectively, or to run your own benchmark test, visit vertica.com. Get connected hp.com/go/getconnected Current HP driver, support, and security alerts delivered directly to your desktop Copyright 01 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice. The only warranties for HP products and services are set forth in the express warranty statements accompanying such products and services. Nothing herein should be construed as constituting an additional warranty. HP shall not be liable for technical or editorial errors or omissions contained herein. AA-089ENW, Created October 01