Solid Information Technology was a leading provider of fast, always-on, and affordable database solutions. Originally founded in 1992 in Helsinki, Finland, Solid had the early vision to provide innovative database solutions that allow fast, failure-proof access to data. Starting with the first product release in 1994, Solid has served the most demanding needs of real-time applications that are deployed in next-generation networks, trading rooms, multimedia network printers, medical devices, fleet management systems, and Point of Sale (POS) solutions, among others. Today, Solid Information Technology is part of the IBM Information Management portfolio and has hundreds of customers in a variety of industries. There are more than 3,000,000 deployments of solidDB worldwide in telecommunications networks, enterprise applications, and embedded software and systems. Market leaders such as Alcatel, Cisco, EMC2, HP, Nokia, and Siemens rely on solidDB for their mission-critical applications.
IBM provides the industry's leading and broadest information management portfolio, delivering high-speed, uninterrupted data access to mission-critical applications.
The introduction of IBM solidDB 6.3 universal cache provides IBM with the ability to accelerate access to IBM DB2 and IBM Informix® Dynamic Server(IDS), as well as Oracle databases, increasing their performance up to ten times. When storing performance-critical data from one of the supported disk-based databases into solidDB's in-memory cache, applications can access data with extreme speed because it is always kept in the computer's memory, rather than on disk. With solidDB Universal Cache, both existing and new applications can now generate data workloads of more than 120,000 transactions per second and safely rely on predictable response times measured in microseconds to support growing numbers of users and data volumes, allowing companies to quickly unlock the business value of their data, which is what Information On Demand is all about.
What does IBM solidDB Universal Cache bring to DB2?
With the availability IBM solidDB 6.3 Universal Cache, you have an industry first in-memory database software to accelerate IBM’s own DB2 and IDS, and Oracle databases. IBM solidDB Universal Cache integrates with relational, disk-based databases to accelerate access to their data. This acceleration can be particularly valuable for performance-critical data frequently accessed by many concurrent users, typical of applications such as ticketing and reservation systems, e-commerce applications, event processing and alerting, online gaming, and software as a service (SaaS) application platforms. Additionally, data from leading disk-based, relational databases can be cached into solidDB Universal Cache to help support peak workloads experienced by such applications as customer service and stock trading.
Performance is enhanced even further because IBM solidDB Universal Cache can be co-located and linked with the application. This means that IBM solidDB runs in the same process as the application. This avoids network latency and the overhead of inter-process communication, as well as context switching in the operating system, further contributing to extreme speed.
This article concentrates on how solidDB Universal Cache integrates with DB2 to accelerate access to data.
Getting started with solidDB Universal Cache
Figure 1 shows the basic architecture and components of an IBM solidDB Universal Cache solution, including the IBM InfoSphere® Change Data Capture (CDC) technology:
Figure 1. IBM solidDB Universal Cache architecture
The key elements of the architecture include:
- solidDB: The front-end database or cache.
- RDBMS: The back-end database used for replication.
- InfoSphere Change Data Capture (CDC): The replication tool that allows you to replicate data between the cache and the RDBMS.
- InfoSphere CDC Access Server: The server that manages the replication processes for the cache and the RDBMS.
- InfoSphere CDC Management Console: A GUI application used to configure, manage, and monitor the replication processes.
Details of the CDC architecture are later explained in the "Configuring the IBM solidDB Universal Cache solution" section.
IBM solidDB Universal Cache 6.3 is a solution that can be configured to incorporate features like High Availability (HotStandby), to enable 24/7 service with transparent fail-over, and Advanced Replication, to accommodate load-balancing or mobile devices that may be disconnected from the system. This article concentrates on a simple setup, initially a single solidDB database as the front-end and a single DB2 database as the back-end, then adding a further front-end database to demonstrate the HA implementation in IBM solidDB.
Prerequisites and installation preparation
Before starting any implementation or testing with solidDB Universal Cache, it is imperative that you plan out exactly what you wish to achieve. There are many configurations available with solidDB Universal Cache, however it is normal to start with a simple solution with bi-directional replication on a few tables. This example is demonstrated in Part 2 of this series.
Please make sure that you have a working DB2 for LUW system, then check the requirements for solidDB and InfoSphere Change Data Capture. Take a look at the following list to find detailed system requirements for each product:
- solidDB system requirements:
- Approximately 64MB of disk space for the code, documentation, and the test database
- At least 40MB of memory for the default configuration, plus space for each table you intend to place in memory
- InfoSphere CDC requirements:
- InfoSphere CDC requires approximately 220MB of disk for each data source
- InfoSphere Access Server requires approximately 250MB of disk
- InfoSphere Management console requires approximately 250MB of disk
- 1024MB of memory recommended
You can download the IBM solidDB Universal Cache software. The trial software has a 90-day licence and includes:
- IBM solidDB
- IBM InfoSphere CDC for IBM solidDB
- IBM InfoSphere CDC for IBM DB2
- IBM InfoSphere CDC Access Server
- IBM InfoSphere CDC Management Console
- IBM solidDB Documentation Package
Installing the IBM solidDB Universal Cache solution
The installation of the IBM solidDB Universal Cache solution is essentially the same, independent of the platform. Part 2 in this series describes the installation of all components and teaches you how to configure the components to demonstrate their use with a DB2 back-end database. This article describes the sequence of installation and configuration in very simple terms. This article assumes that the back-end database, in this case DB2, is already installed and in-use.
Before you begin
Before you begin the installation, check that you have:
- All the required software packages.
- Sufficient privileges or authority to install the software.
Figure 2. IBM solidDB Universal Cache installation sequence
Your first step is the installation of IBM solidDB which, in most cases, simply requires the administrator to run the installation program and set the CLASSPATH environment variable so that it includes the IBM solidDB JDBC driver .jar file. If you want to access the English version of the documentation in PDF format, you can unzip the separate documentation package into the manuals subdirectory.
Once the installation is complete, it is useful to start IBM solidDB to ensure that the installation was successful. To start IBM solidDB:
- For Linux/UNIX: Enter the command
solid -fat the command prompt to force the server to run in the foreground.
- For Windows: Click on the icon labeled IBM solidDB Server from the Start Menu.
You might want to set up a link to the IBM solidDB 6.3 and IBM solidDB Universal Cache 6.3 Information Center (see Resources for the link). Here you can find information describing how to use the IBM solidDB family of products and features. The installation documentation for InfoSphere CDC Access Server and Management Console can be found in the download package (AS_MC_InstallationGuide.zip). The package only contains the installation details for the Access Server and Management Console; all the other InfoSphere documentation is included in the software download packages and will be installed during the software installation procedure.
The second step is the installation of the InfoSphere CDC component for solidDB. This is the replication element that enables you to replicate data to and from solidDB. The package contains only the installable image; all documentation is in the documentation package.
At the end of the installation, if you have solidDB up and running, you can launch the CDC for solidDB Configuration Tool to verify the installation was successful. You can create a default instance using the standard defaults, for the users and port numbers.
NOTE: Verification of installation would normally be completed after the installation of CDC for IBM solidDB and the back-end database. You must have the details for the users and the port numbers before you can create the instance.
The third step is the installation of the InfoSphere CDC component for the back-end database. This may well be in place if you are already using InfoSphere CDC for replication. If you have had to install InfoSphere CDC for the back-end, at the end of the installation, launch the CDC for DB2 Configuration Tool to verify the installation was successful. You can create a default instance using the standard defaults, for the users and port numbers.
If you use the code with the download package, you will have 14 days to evaluate the product. If you require an extended test, please contact your local IBM representative for details.
If you already have a replication solution, using SQL or Q-Replication, installed for the back-end database, it is necessary to ensure that there are no conflicts.
Your fourth step is the installation of the InfoSphere CDC Access Server. This is the workstation that controls the processes for the replication and the communications with the Management Console. If you are using a firewall or any other security mechanism, you may need to confirm static ports for the replication and communication processes. You can do this by editing the dmaccess.vmargs file.
The fifth step is the installation of the InfoSphere CDC Management Console. The console is the GUI that lets you manage your replication environment.
Configuring the IBM solidDB Universal Cache solution
The configuration of the IBM solidDB Universal Cache solution requires the implementation of a replication subscription between the solidDB Universal Cache and the back-end database. This section provides an overview of the configuration process to implement this.
Figure 3. IBM InfoSphere CDC architecture
The key components of the architecture include:
- Datastore (also known as Replication Agent) - a process on source or target servers that accepts the requests from the Access Server and communicates with the Replication Engine to manage replication activity.
- Replication Engine - a process that sends or receives replicated data.
- Access Manager - an integrated component of the Management Console. You will use the Management Console to manage datastores and user accounts.
Before you begin
Before you start to configure your solution, check that you have:
- Your databases up and running.
- Sufficient privileges or authority to configure your environment.
- The details for users and port numbers of the CDC instances.
- Defined your desired replication model.
The first step is the creation of an instance of InfoSphere CDC for IBM solidDB. If you created the default instances in Step 2 of the "Installing the IBM solidDB Universal Cache solution" section, this can be used in most cases. Otherwise you will need to launch the configuration tool (dmconfigurets, which can be found in the bin subdirectory for InfoSphere CDC).
Your second step is the creation of an instance of InfoSphere CDC for the back-end database. If you created the default instances in Step 3 of the "Installing the IBM solidDB Universal Cache solution" section, this can be used in most cases. Otherwise you will need to launch the configuration tool.
The third step is logging in to the Management Console. When you log in to the Management Console, you are connecting to the Access Server. The initial login uses the System Administrator user that was created during the installation. The Systems Administrator has access to all available operations in the Management Console. It can be used to assign permissions to specific users to enable delegation of duties. All future steps are performed within the Management Console and should be based on your replication model.
Your fourth step is the creation of new datastores for solidDB and the back-end database. When adding a new datastore, you:
- Must specify information about the database and provide database connection parameters so that users can connect to the datastore.
- Need to assign both datastores to the same user(s).
- Can (optionally) set the default connection behavior for the Access Server. The administrator can define the default port, specify the outbound ports for the Access Server, and set the system to automatically connect to the datasources when logging in.
The fifth step is the definition of replication model between solidDB and the back-end database. This example has a simple bi-directional model between solidDB and DB2. When adding this new replication model, you:
- Must define the tables, to be replicated, in the source and target databases.
- Set up subscriptions, from solidDB to DB2 and DB2 to solidDB.
- Map the tables for replication in both subscriptions. For each mapping, set the conflict detection and resolution.
- Start the replication.
IBM solidDB Universal Cache is relational in-memory database software that delivers extreme speed, performing up to ten times faster than conventional databases. It uses the familiar SQL language to allow applications to achieve throughput of tens of thousands of transactions per second with response times measured in microseconds, also providing applications extreme data availability with sub-second failover. It can be deployed as a cache for IBM DB2, IDS, or Oracle RDBMS to deliver performance-critical data with extreme speed.
This article has presented the IBM solidDB Universal Cache solution and explained its component parts. It has outlined the basic system requirements and the architecture of a solution. This article has described how to get started with the IBM solidDB Universal Cache solution and a DB2 LUW back-end database.
Stay tuned for Part 2, which takes you through the steps to implement a sample solution, complete with a sample application and bidirectional replication with a DB2 back-end database.
|IBM solidDB Universal Cache 6.3 Data Sheet||Universal_Cache_6.3.pdf||150KB|
- Introduction to IBM solidDB Universal Cache 6.3
- "Introduction to IBM solidDB Universal Cache 6.3, Part 2: IBM solidDB Universal Cache setup" (developerWorks, May 2009): Set up IBM solidDB Universal Cache with IBM DB2 for Linux, UNIX, and Windows to accelerate access to data.
- IBM solidDB 6.3 and IBM solidDB Universal Cache 6.3 Information Center: Find information describing how to use the IBM solidDB family of products and features.
- "IBM Announces Industry's First Universal Database Cache" (IBM, October 2008): Read the press release for IBM solidDB Universal Cache 6.3.
- IBM solidDB Services, Get more information about solidDB technology, which provides a state-of-the-art methodology for effective, reliable services.
- developerWorks Information Management zone: Learn more about Information Management. Find technical documentation, how-to articles, education, downloads, product information, and more.
- Stay current with developerWorks technical events and webcasts.
- Technology bookstore: Browse for books on these and other technical topics.
Get products and technologies
- IBM solidDB Universal Cache: Download a trial version of IBM solidDB Universal Cache.
- Build your next development project with IBM trial software, available for download directly from developerWorks.
- IBM PartnerWorld®: Business Partners and ISVs can utilise PartnerWorld to get up to speed with detailed product information, software downloads, and service options.
- Participate in the discussion forum.
- Participate in developerWorks blogs and get involved in the developerWorks community.