Making a Happy Marriage of WebSphere & TIBCO Infrastructures (Part 4 of 6)

TIBCO Metrics

(Fourth of a six-part blog series that describes how a team of IT pros and managers at one of the world’s largest global banks accommodated a bank acquisition and mastered a complex messaging environment.)

A major driver of MegaBank’s decision to standardize on Nastel AutoPilot was its ability to monitor all of its middleware systems. With AutoPilot’s TIBCO EMS plug-in, the IT team could easily monitor TIBCO EMS servers and components including: queues, topics, consumers, producers from a single, consolidated vantage point.

The Backstory on MegaBank’s TIBCO EMS Monitoring Solution

At MegaBank (or any other customer utilizing TIBCO), the TIBCO EMS (Enterprise Message Server) Plug-in Expert (a “plug-in expert” is a collection of polices and instrumentation to automate the monitoring of a specific technology, e.g., JMX) is modularized based on EMS servers. The AutoPilot TIBCO EMS Plug-in retrieves information such as queues, topics, consumers, producers and more, plus their associated components. For example: queue expert, topic expert, consumer and producer experts, etc.

A customer’s specific requirements for performance monitoring activities determine how plug-in experts are utilized—individually or in any combination of multiple deployments. AutoPilot’s TIBCO EMS Plug-in Expert automatically measures and collects EMS performance metrics, periodically gathering information about TIBCO EMS servers and their components on a sampling interval defined by an IT middleware team. Some key performance metrics collected includes:

• Inbound
• Outbound
• Acknowledged and pending message volumes
• Message rates per second
• Average response times
• Message tracking
• Age of oldest messages

These metrics are collected for servers, queues, routes, groups, consumers, producers, transports, and durables. (Creating a “durable” data subscription means it is stored on the server and subscribers can receive messages even if it was inactive when the message was originally delivered. Otherwise, if messages are published when the subscriber is not available, the subscriber does not receive those messages and events.) AutoPilot’s powerful Complex Event Processing (CEP) engine uses these metrics in processing rules (a.k.a. policies) an IT middleware team defines for predictions, correlations, validation, automation, notifications, logging and alerts.

AutoPilot’s CEP engine routinely combines metrics and data from many diverse systems within a customer’s infrastructure. In the case of MegaBank, one of the most important features for the firm was the fact that the AutoPilot TIBCO/EMS plug-in can monitor multiple TIBCO EMS servers. (Actually, there’s no limit to the number of TIBCO EMS servers that can be simultaneously monitored.)

The AutoPilot CEP engine correlates the EMS metrics with measurements from other infrastructure components such as DataPower, Solace, WebSphere MQ, Message Broker, ESB, and other supported technologies, dynamically creating events and alerts for IT staff as needed.

Lifting the Hood on MegaBank’s TIBCO Rendezvous Monitoring Solution

An additional requirement MegaBank had for AutoPilot was performance monitoring for TIBCO Rendezvous (RV) and visibility into TIBCO multicast services. Using AutoPilot’s TIBCO RV Plug-in expert, the IT middleware team could monitor TIBCO RV performance statistics. The plug-in expert automatically collected data on the TIBCO RV server and its components (services, routers and transports).

Specifically, the plug-in expert collected TIBCO RV statistics by subscribing to and processing its advisory messages. The TIBCO RV Plug-in Expert analyzed collected data and provided performance analytics for:

• TIBCO RV Server Info
• Client Transport
• Service
• Router Info
• Security Info

MegaBank needed this functionality in both agent and agent-free deployments. (In some environments, the bank was unwilling to deploy anything additional to their production servers and therefore required an agent-free implementation). The TIBCO RV Plug-in Expert utilized metrics provided by the TIBCO RV server; this data is used in AutoPilot business views for validation, automation, notification, logging and alerts.

An AutoPilot Business View showing the health of the TIBCO RV Server environment displayed RV status, including:

• Server health state
• Inbound and outbound multicast lost messages
• Inbound and outbound point-to-point lost messages
• Number of clients using a particular service
• Number of hosts using a particular service
• Rate of inbound and outbound traffic
• Amount of inbound and outbound data
• Number of missed inbound messages

AutoPilot provided MegaBank’s IT team with the visibility and performance metrics they needed to properly monitor their TIBCO messaging components—just as it did for WebSphere-based elements.

In next week’s installment of Making a Happy Marriage of WebSphere & TIBCO Infrastructures – Part 5, I introduce the concept of message tracking and the very tangible benefits this capability delivered to MegaBank.