There are dozens upon dozens of research analyst reports available that extol the virtues of cloud computing. According to these reports, the cloud computing model offers your organization greater flexibility and efficiency than your current computer systems designs — and a better return-on-investment (ROI) thanks to the virtualization and automatic provisioning of computing resources (systems, storage, and networks).
Now think back 5 to 7 years. Greater flexibility, improved efficiency, and better ROI are the exact benefits that grid computing was supposed to deliver. But the problem with grid computing was that the management of applications across distributed resources was too complex (and also too difficult to secure) — hence, grid applications were confined behind-the-firewall where they could be more easily managed and contained.
The lesson that we all should have learned from grid computing is that the most important element in making a distributed computing model such as cloud computing work is application management. We have to know where our resources are; what service level those resources provide (such as security, resiliency, and scalability characteristics) — and we have to understand how our applications and databases behave when using those distributed cloud resources.
When it comes to cloud computing, Clabby Analytics is particularly focused on application management. More specifically, we follow the business transaction management (BTM) and application performance management (APM) disciplines. What we have found in our research is that there are dozens of products in these categories can be used to track application flows across multiple systems types — and can be used to ensure that service level requirements are met. There are some technical differences in the approaches that APM/BTM vendors use to monitor and manage applications and transactions within a distributed computing environment — but, for the most part, the main differentiators can be found in the depth and breadth of each respective vendor’s product offering.
In this Research Report, Clabby Analytics defines APM and BTM, and outlines the types of problems that these tools solve. We look closely at the product offerings of an emerging APM/BTM vendor — Nastel Technologies. We first reviewed Nastel in June of 2010, and at that time, we found that Nastel had a complete, broad and deep APM environment that effectively tracks and monitors applications and transaction flows within a cloud. After a recent update, Clabby Analytics believes that Nastel has made significant progress in the last six months — announcing new partnerships, expanding cloud capabilities, and working closely with large customers in financial services, telecommunications, healthcare and retail to develop new use cases for their APM technology, largely being enabled by Nastel’s sophisticated complex event processing (CEP) engine.
A Closer Look at APM and BTM
The BTM market is an offshoot of the application performance management market. As defined by Wikipedia, APM “refers to the discipline within systems management that focuses on monitoring and managing the performance and service availability of software applications.” Wikipedia describes BTM as “an approach to managing IT from a business transaction perspective. BTM aims to guarantee service quality for users conducting business transactions while simultaneously optimizing the IT applications and infrastructure across which those transactions execute”. At the heart of BTM lies the ability to capture and to track all transactions, across all IT tiers, automatically and continuously. And then there is Business Transaction Performance (BTP), defined by Wikipedia as “methodology providing for comprehensive visibility into complex, multi-tier business transactions. BTP provides real-time discovery of transaction topology, behavior and performance; visualization, analysis, automation, reporting and process improvement.” Since applications are comprised of transactions, these disciplines often go hand-in-hand — with many enterprises using both BTM/BTP and APM, or with businesses using APM tools that encompass some of what we associate with BTM or BTP and vice versa.
BTM and BTP have arisen largely out of necessity, since today’s applications span multiple computing tiers and are run in distributed environments. Cloud computing exacerbates the need for APM and BTM software products.
For an APM tool to be effective, it either must be combined with a BTM or BTP tool or have the capability to look at applications from a distributed, transactional point-of-view.
With the market increasingly moving to service level agreements (SLA’s) both in public cloud environments and private clouds, performance monitoring and management must be proactive. And in order to address performance issues before SLA’s are breached, customers must have visibility into both application performance and transaction status. By providing proactive monitoring that can identify problems and pinpoint the source, using APM and BTM tools can significantly reduce support costs and improve end-user experience. By correlating IT issues with business impact as well as identifying performance trends and patterns, these tools can be used to improve business processes, reduce time-to-market for new applications, boost productivity, and improve customer service.
Anatomy of a Transaction
A transaction can be defined as an exchange of information between any two parties. A software transaction is an atomic (that, is if one component of the transaction fails, the entire transaction fails) transaction. Atomic web services transactions used in distributed applications are managed by the protocol WS-Atomic Transaction (WS-AT) which does not allow a partial transaction to occur – it is all or nothing. A classic transaction, such as a transfer of funds or an on-line purchase, is made up of many atomic transactions. An atomic transaction will either commit (complete) or abort (fail). This characteristic of atomic transactions ensures that partial updates to a database do not occur – or as in our funds transfer example- that funds don’t come out of one account and not reach the other account. In this case if there were a failure before the funds were deposited in the target account, they would be rerouted back to the origin account. In order to manage transactions, software must be able to reassemble the various atomic elements that make up a transaction and relate them to the entire business transaction flow ─ a funds transfer, for example. This capability gives IT administrators insight into:
- how long a business transaction takes to complete;
- measure the time taken for each atomic transaction that makes up the business transaction; look at the path of a business transaction; and,
- determine if the transaction is failed or complete
Tracking application and transaction faults in a distributed systems environment or a cloud is complex. It involves following applications and transactions as they hop from server to server making database calls and performing reads and writes along the way. As you can see from our example, if a funds transfer transaction is not successful, it is possible that the failure could have occurred anywhere along the way.
It is crucial that IT administrators are able to identify the location and cause of the failure so that quick, corrective action can be taken. Further, tracking transactions helps to pinpoint performance issues, identify trends and measure the usage of IT resources. And with the monitoring and tracking capabilities of APM, BTP and BTM tools, problems can, in many cases, be identified and fixed before transactions failures occur.
About Nastel Technologies
Nastel, founded in 1994, is a privately held maker of APM/BTM products, headquartered in New York (with offices in the U.S., Europe and Mexico). Nastel’s partners include IBM, HP, Microsoft, Oracle, Sun, RedHat, Coradiant, Solace Systems and Kratos Networks.
Nastel focuses on providing APM/BTM monitoring environments for distributed systems and mainframes ─ and the company has particular expertise in monitoring the performance of transactions and middleware messaging running across a wide range of computing environments.
Nastel also has a comprehensive set of service offerings including consulting, training, implementation, and technical support.
Nastel has a very strong presence in the financial services market (as one would expect since financial services is heavily transaction oriented) with a worldwide customer base including Citi, JP Morgan, Credit Suisse, Bloomberg, Deutsche Bundesbank, HSBC and many others. Other focus industries include healthcare and insurance, retail, manufacturing, and telecommunications with customers including Amica, Blue Cross/Blue Shield, Northwestern Mutual, AT&T, Staples, Best Buy, Sky Telecom, Debenham’s, Dell, ExxonMobil and Sony Ericsson.
Nastel’s AutoPilot® Suite is an APM management environment; Nastel’s TransactionWorks™ is a business transaction management environment that can work in conjunction with AutoPilot. Nastel AutoPilot for Cloud was specifically designed to address cloud-based application availability and performance issues.
Nastel AutoPilot Suite – A Closer Look
Nastel AutoPilot M6 product suite monitors and reports on transactions, application performance, middleware, and business connectivity in both distributed and mainframe environments. AutoPilot combines the benefits of BTM and APM, by enabling an end-to-end look at transaction level data as well as the ability to drill down in a particular layer to diagnose a problem.
At the core of AutoPilot is an embedded Complex Event Processing (CEP) rules engine that processes millions of rules, metrics and messages, correlating operational and transactional data and placing it in a business context for “360 degree situational awareness”. This engine provides data that can reduce the length of application degradation and outages (MTTR) and supports proactive problem identification. What is also significant is that AutoPilot works with existing monitoring tools, and can collect and correlate virtually any type of metrics including business key performance indicators (KPIs), information from environmental sensors, “smart” devices, RSS and news feeds, and email messages. High speed, high volume rules–based processing ensures high availability and optimal performance for mission-critical applications.
The Nastel AutoPilot M6 suite is designed to monitor and manage distributed IT services and user applications that use application servers, Enterprise Service Bus (ESB), messaging middleware, workflow engines, brokers, and Service Oriented Architecture (SOA).
Nastel’s distributed implementation of CEP combined with a policy engine, statistical baselines, and a wizard interface enables customers to quickly and easily identify whether operations are normal or abnormal, delivering governance for application performance.
AutoPilot was built using a service-oriented architecture (SOA) and highly available GRID technology. AutoPilot’s SOA architecture provides easy integration with other management tools and monitoring for a wide range of technologies: J2EE, .NET, SNMP, TIBCO RV and EMS, DB2, Oracle, SQL Server and IBM WebSphere, and most recentlySiebel, SAP and Business Process Execution Language (BPEL). The GRID architecture supports hot failover and linear scalability.
Other features include:
- Integrates with data from external sources – RSS and news feeds and email
- Can broadcast real-time alerts over Twitter
- Allows users to define policies based on specified rules and conditions
- Integrated Business Activity (BAM) support
- Customizable dashboard
- Granular security, authentication and access control
Nastel’s TransactionWorks ─ A Closer Look
TransactionWorks is part of the AutoPilot suite and specifically drives BTP including auto-discovery of transactions and real-time tracking of application performance for a range of applications including payment, trading, claims processing, retail order management and patient data transmission.
TransactionWorks can be used along with AutoPilot’s predictive capabilities to track application/transaction behaviors and flows in order to answer industry specific questions such as:
- Financial Services
- Health Care
─ What happened to my order? Where did it get bogged down? The database? The application? A middleware layer? Did it not get credit approval?
─ Has my order return been processed? My account credited?
- ─ Did the trade fail? What is the root cause of a failed trade?
─ Does the trade violate the 3-day settlement compliance mandated by the SEC?
- ─ Did an Electronic Medical Record (EMR) request go through?
─ Have compliance regulations been met? Has patient been booked for surgery?
─ What is the status of a patient’s healthcare claim?
TransactionWorks tracks transaction compliance including number of transaction, in-flight transactions, violations, violation rate, failures, failures rate, maximum and average duration. TransactionWorks also provides visibility into messaging middleware, which provides information that can piece IT transactions together so that the complete transaction (request and all subsequent messages and responses) are captured and connected for a complete view of the whole business process transaction).
Other features include:
- Root-cause analysis with drill-down to identify performance bottlenecks
- User defined SLA’s
- Records activity in database to support audit requirements
- Transaction analysis to identify trends and provide input to capacity planning
- Business Service Views
- Track messages exchanged between applications
- Complete audit trail
- Business rule to detect “business normal” vs. “business abnormal”
Figure 2 shows how Nastel’s AutoPilot (in conjunction with TransactionWorks) monitors J2EE, messaging middleware, and CICS mainframe environments, capturing transaction activities in real-time. Visibility into each step of the transaction allows a “replay” of the trade should something goes wrong. Using the trade ID, stored in Websphere Message Queue (WMQ) middleware and also in each phase of the transaction, the trade can be pieced together and presented as a complete business transaction — making it possible to discover what problem occurred in the transaction process flow ─ and why it happened.
Nastel’s AutoPilot for Cloud
AutoPilot for Cloud is a preconfigured virtual appliance with AutoPilot M6, TransactionWorks designed for private and hybrid (mix of public and private) cloud environments, AutoPilot for Middleware and prebuilt governance policies. The virtual appliance includes application, transaction and middleware performance management in a single VMware image running a Linux Guest OS.
Included is built-in, cloud-ready grid support and CEP situational Analysis. This preconfigured, prepackaged offering promotes rapid deployment in the cloud, quick provisioning, high-availability, and ease of management for mission critical cloud computing environments.
- Easy failover for cloud based applications that demand 24x7x365 availability and on-demand capacity as load increases.
- Self-Service with role-based views for ease of use
- No service interruption when performing replacement and maintenance on hardware platforms running AutoPilot
- Transparent movement of the appliance from User Acceptance Testing (UAT) to production without concern for hardware issues
- CEP Situational Analysis with business rules prevents recurring problems
- Twitter interface for ease of communication and management
Customer use cases
In Clabby Analytics most recent briefing, Nastel shared stories of how customers are using Nastel AutoPilot to derive value. Here are a few examples: A large retail chain is using Nastel AutoPilot to monitor pricing in their retail stores. Before implementing AutoPilot, prices were inconsistent across stores. Some retail stores weren’t up-to-date with current corporate pricing, and had systems that were unable to alter prices in real-time as they changed. But it was difficult to identify which stores were affected. By using Nastel AutoPilot to monitor systems in all stores, this retail chain was able to maintain transactional consistency, ensuring accurate and uniform pricing throughout the chain.
In a large manufacturing company, a problem with the monitoring system triggered a cascading failure that brought down an entire assembly line. After installing Nastel AutoPilot, the IT staff was able to see every server and any point of failure in the manufacturing process control systems, from a consolidated dashboard view. Nastel AutoPilot identified the problem — and using event data, correlation and business rules — created a solution that will prevent the problem from recurring. Traders in a large financial services company, in some situations, were unable to get a dial tone when attempting to make a phone call. In order to prevent this, AutoPilot’s monitors the phone system, looking for events that trigger this situation. Data is collected and correlated to proactively identify potential “no dial tone” and calls are automatically rerouted.
In October 2010, Nastel announced a partnership with Coradiant, a leader in end-user experience monitoring for web-based applications. Nastel AutoPilot integrated with Coradiant TrueSight will combine end-user experience monitoring with “deep dive” application and transaction performance monitoring. As a result, customers will gain insight into not only detailed application performance and proactive problem detection but also will have the ability to see how these issues are impacting the end-user.
IT personnel will have a dashboard view of information that compares the performance of each transaction to Service Level Agreements (SLAs) in order to ensure compliance. Operations will be able to view the real-time performance of end-users by geography, user group, transaction type, time-of-day and other criteria.
Another recently announced partnership is with Kratos Networks, a network monitoring and management company. As part of the partnership, Nastel AutoPilot will be integrated with Kratos NeuralStar (for enterprises and distributed networks) and dopplerVUE (for networks of any size) network event management. The combined solution will provide application aware network monitoring. Network data will be collected and fed into the AutoPilot CEP engine, providing additional data that can be correlated and used to better predict performance issues and potential transaction failures. Automated responses are based on policy driven tasks and events. This capability will span multiple networks and IT systems and can monitor and manage any network attached device.
This is particularly important in public and hybrid cloud environments, where network monitoring can collect “deep-dive” application data (including the ability to detect individual applications) which might otherwise be inaccessible. This deep analysis with specific application and transaction level data improves incident response times and customer service while reducing the support resources required to identify and fix problems. Understanding the interactions and interdependencies between applications and networks enables speedier resolution and improved productivity.
On November 16, 2010, Nastel announced a partnership with Solace Systems, a provider of messaging middleware appliances with a large presence in the financial services market.
AutoPilot for Solace plug-ins support all major Solace appliances for real-time monitoring of detailed performance information, events generated by the appliances, and the performance of VPNs and the Solace message cache. The solution leverages the Nastel CEP engine to identify problems before impacting SLAs, and in real time, captures, stores and reports on performance trends and statistics, and provides customizable business views and proactive alerts.
Nastel’s support for a blend of APM, BTM and BTP capabilities has earned them a spot as a “Visionary” for APM in Gartner’s Magic Quadrant, supporting all five dimensions of APM, with “deep expertise in WMQ and Java EE technical sophistication and code excellence”. And Forrester Research recently recognized Nastel as a vendor providing one of the best tools to build a closed-loop automation solution for application management in a December 2010 report titled “Evaluating Innovative I&O Solutions: Converged Application Performance Management – Nastel Uses Complex Event Processing To Manage Application Performance.” For an overview of the BTM marketplace and a description of major competitors in this market place, please see Clabby Analytics’ free BTM market overview report entitled “How New Business Transaction Management Tools Are Making It Possible to Address Transaction Performance Tuning and Capacity Planning Issues in the Cloud” that can be found at ttp://www.clabbyanalytics.com/uploads/BTMRevUpdate.pdf.
From our perspective, Nastel’s direct competitors include CA Technologies (Wily Introscope and Customer Experience Manager), IBM (various Tivoli products) and HP’s TransactionVision. Other competitors include Optier, Dynatrace, Correlsense, and Precise.
To Clabby Analytics, the design point of the future in information infrastructure is “cloud computing”. Cloud designs improve resource utilization (leading to greatly improved return-on-investment) ─ while also leading to dramatically reduced operational costs (certain operations can be outsourced quite effectively, while those operations run inside an enterprise firewall can be better managed using advanced management software). Nastel is a provider of the type of advanced management software that will lead to lower cloud operating costs (by simplifying application control, monitoring, and troubleshooting).
What we look for in APM/BTM management environments such as the AutoPilot/TransactionWorks environment offered by Nastel are four things:
1.the ability to automatically track transaction flows across distributed systems environments (without having to manually “sniff” for problems);
2.the ability to see business process flows;
3.an intuitive interface; and,
4.integration of management functions.
To us, Nastel offers a graphically-oriented, intuitive management environment that allows application/transaction behaviors to be tracked automatically with a minimum of human-related manual labor involved. Further, we like the ability to see graphical representations of process flows along with representations of where a transaction or application may be failing or where tuning needs to take place. Nastel’s CEP engine is a strong differentiator because it can collect and correlate so many useful metrics from a wide variety of sources, and analyze this data to link IT problems to business impact and, based on policies and rules, provide an automated response.
And in areas where Nastel doesn’t have a specific solution (end-user experience monitoring, for example), they have aggressively pursued best-of-breed partners (like Coradiant) to augment their offering, while maintaining a focus on improving their core technology.
Telephone: 001 (207) 846-6662
Clabby Analytics is an independent technology research and analysis organization that specializes in information infrastructure and business process integration/management. Other research and analysis conducted by Clabby Analytics can be found at: www.ClabbyAnalytics.com. For enterprise IT managers looking to deploy a computing cloud, Nastel’s APM/BTM offerings represent a way to manage applications as they flow through the amorphous cloud. APM/BTM tools and cloud-enabled offerings, such as those provided by Nastel, are vitally important when it comes to ensuring high levels of service for applications within the cloud ─ as well as to reducing operational and support costs related to the management of the cloud.