Tech2Tech

Applied Solution 4

High availability, low worries

Teradata Multi-System Manager keeps mission-critical systems running at peak performance.

Figure 1: Dual-system environment

Click to enlarge

The customer is king. And when the king demands up-to-date information from vendors, suppliers and merchants 24x7, those groups had better be able to deliver. To accommodate such re­quests, organizations rely on mission-critical applications, which can vary from tracking shipments to making real-time offers to customers as they interact at the various touch points.

Underscoring their importance to organizations, these applications often have very rigorous require­ments for high availability and disaster recovery. To accommodate, an application typically executes in an environment that has at least two geographically separated systems. High availability can be achieved with cluster techniques at a single site, but true disaster recovery protection requires the ability to recover at a geographically separate site. Having two systems also provides the ability to dramatically reduce downtime for maintenance or upgrades as well as the potential to balance workload to meet peak demands.

Teradata’s solution for supporting mission-critical applications in a dual-system environment consists of several products that interact with at least two Teradata systems that can be of different sizes and different hardware. (See figure 1.) Data synchronization is performed via:

  • Dual loads for high-volume, low-latency tables
  • Replication for real-time, low-volume, low-latency changes
  • Teradata Data Mover for remaining tables

While Teradata Query Director provides the ability to route application workloads between the two systems, Teradata Multi-System Manager provides the centralized monitoring and control of the dual-system environment and is responsible for keeping an envi­ronment up and running at peak performance.

A controlled environment

A reality of these new mission-critical environments is increased complexity within the computing infrastructure. Areas that need to be tightly controlled are:

  • Availability of components such as extract, transform and load (ETL); business intelligence (BI); and database servers
  • Data synchronization between databases and tables
  • Application failover, switchover and workload balancing

Teradata Multi-System Manager greatly simplifies the manage­ment of a mission-critical Teradata environment. It’s an event-driven, exception-based application designed to continually monitor and manage multiple Teradata ecosystems. Each ecosystem includes at least one Teradata Database system along with the servers and pro­cesses that load or retrieve data from the databases.

Teradata Multi-System Manager is a packaged solution consisting of:

  • Managed servers (single or redundant) running the Teradata Multi-System Manager application
  • Java-based monitoring agents for Linux, Windows, Solaris, etc.
  • Teradata Viewpoint software running on a managed server that enables Teradata Multi-System Manager portlets

Its key features include the ability to:

  • Monitor the state (active/out of service) and condition (speed of response) of applications
  • Switch application workloads between systems by changing rout­ing rules in response to a state change
  • Compare all events with thresholds and generate alerts or actions when anomalies are detected

The administrator's role

Teradata Multi-System Manager’s role

Teradata Multi-System Manager leverages a sophisticated data model to easily manage complex relationships among the com­ponents of a multi-system environment and maintain the rules for event-driven exceptions. This data model was engineered from the ground up to meet these needs.

To carry out its mission, Teradata Multi-System Manager:

  • Continuously analyzes events received from monitor agents (i.e. “file system full,” “component heartbeat failed”) and processes such as load utilities, data health, replication, Teradata Data Mover and Teradata Query Director for various sources in the ecosystems
  • Compares event values with defined thresholds to determine condition of applications, components, processes and tables
  • Generates automatic actions in response to condition exceptions or state changes of applications, components, processes and tables by:
    • Contacting an administrator or application via e-mail
    • Performing remote command execution or Simple Network Management Protocol (SNMP) trap
    • Providing a visual display on Teradata Viewpoint portlets

Ease with Teradata Viewpoint

Figure 2: Teradata Multi-System Manager portlets

Peace of mind

With the demands placed on modern companies to provide their customers with up-to-date information, the mission-critical applications that provide this data must meet rigorous standards for high availability. And when the path to high avail­ability includes a dual-system environment, Teradata Multi-System Manager provides the tools to successfully control the ecosystem. The king can have his data—and the providing organization can have its peace of mind.


Your Comment:
  
Your Rating:

Comments
EMC Q3-2014