DB2 physical database design for the OLTP environment: reliability, availability, and scalability

To keep up with the ever-increasing and increasingly competitive market, your enterprise architecture must be flexible enough to grow along with future strategic needs, and ensures business continuity during planned and unplanned shutdown. For 24x7 mission-critical enterprises, an hourly downtime can be converted into millions of dollars in revenue loss.

To keep up with the ever-increasing and increasingly competitive market, your enterprise architecture must be flexible enough to grow along with future strategic needs, and ensures business continuity during planned and unplanned shutdown. For 24x7 mission-critical enterprises, an hourly downtime can be converted into millions of dollars in revenue loss.

To keep up with the ever-increasing and increasingly competitive market, your enterprise architecture must be flexible enough to grow along with future strategic needs, and ensures business continuity during planned and unplanned shutdown. For 24x7 mission-critical enterprises, an hourly downtime can be converted into millions of dollars in revenue loss, not to mention damage to business reputation and potential customer loss. Global enterprises operate across time zones to provide around-the-clock business services. The scheduled maintenance window for system maintenance and upgrade no longer exists. Distributed enterprises must be able to provide similar services in each geographic location and avoid network faults or transmission time.

Reliability, availability, and scalability a solution implements some or all aspects of reliability, availability, and scalability (RAS) by implementing the following solutions. Based on the capabilities provided by each solution in all three aspects, these capabilities are listed in the worst to best order:

Shared disk Cluster

It provides high availability for node faults in the cluster. This solution only provides high availability and does not provide scalability, disaster recovery, or protection against disk damages.

Disk Mirroring Technology

Many solutions provide the commercial Disk Mirroring Technology to achieve high scalability or disaster recovery through the shared disk cluster solution. However, these solutions do not provide you with full protection when the disk is damaged. If the source disk is damaged, the damaged data is also transmitted to the target. In addition, this solution does not provide the instantaneous failover function, which is critical to 24x7 businesses.

DB2 High Availability Disaster Recovery features

This is a low-cost and easy-to-manage replication solution. It provides high availability and disaster recovery solutions for some and all site faults. It also provides transient failover.

DB2 pureScale®Features

This is a shared disk architecture that enables enterprises to dynamically and transparently expand OLTP clusters as needed. It provides unlimited capacity, reliability, and continuous availability.

Partition Environment

The partitioned database environment is a non-shared architecture that enables the Database Manager to expand to hundreds of TB of data and hundreds of CPUs across multiple database partitions to form a single large database. These partitions can be located in an internal, cross-host, or combined environment. Database Data is distributed across multiple database partitions, providing huge scalability and workload concurrency across these partitions.

Typical OLTP workloads run for a short period of time. The transactions they run only access some random rows in the table. The partitioned database environment is more suitable for data warehouses and smart business workloads, because the communication between instances occurs in each transaction.

RAS solutions suitable for your business

Determine an appropriate RAS solution for your business. You must first define your high availability and disaster recovery goals. To help determine these goals, you need to analyze whether your current enterprise has a ready infrastructure to provide RAS. In order to find a suitable solution, you need to answer the following questions to understand the business impact of downtime.

What are the necessary and important RAS aspects for your business?
What measures have been put in place to mitigate the risk of service interruption?
When the business infrastructure is shut down due to planned or unplanned interruptions:

-What are your business requirements and customer service level agreements?
-What is the impact on your business and customers? Revenue, credibility, future sales, current and potential Customer losses?
-What is an acceptable recovery window in the event of a disaster?
-How long does it take to bring the existing infrastructure back online?

The following sections further discuss two DB2 solutions that provide all three aspects of RAS.

DB2 High Availability Disaster Recovery features

DB2 High Availability Disaster Recovery (HADR) is an easy-to-use Data Replication feature that provides High Availability and Disaster Recovery solutions for partial and complete site faults. HADR copies data changes from the source database (called the master database) to the target database (called the backup database ). Each database uses its own storage. The standby database always applies the transaction logs received from the primary database to maintain synchronization with the primary database. During planned or unplanned shutdown, the Standby database can instantly fail over to the service client. Any cluster server software can be used for automatic failover.

Failover is transparent to clients when combined with automatic client re-routing (ACR) or cluster Servers configured using virtual IP addresses. Tivoli System Automation for Multiplatforms (sa mp) software is a recommended Cluster Server for HADR failover Automation because it is tightly integrated through the DB2 high availability (HA) interface.

An example of the HADR environment is displayed:

Figure 6. HADR Environment

When is HADR suitable for your business? When your business has the following requirements, HADR is suitable for your business:

 A low-cost and simple solution for easy setup and Management of high availability and disaster recovery.
Slave copies the entire database.
 Instantaneous failover of planned and unplanned interruptions.
The slave node uninstalls the read-only workload to the slave database to free the master database from reading or writing workload for key services.
 Provides high availability and disaster recovery solutions for all platforms on the product hardware.
 Does not require dynamic horizontal scaling.