How is BTM different from Application Performance Management (APM) solutions?
Generally speaking, APM solutions combine end-user experience monitoring tools (EUM) with infrastructure-specific diagnostic tools (deep dive tools). The end-user monitoring tools work continuously, and trigger an alarm when performance falls below set levels. At that point, deep dive tools are invoked and start monitoring the data center. Sampling and correlation techniques are used to put all of this information together and display it on a dashboard.
APM tools are an important part of the diagnostic and problem solving arsenal. They provide large amounts of very detailed information about specific infrastructure tiers. Because of the large volumes of data, they typically do not run continuously in the production environment – instead, they are designed to help solve a problem once it has been detected.
BTM is designed to continuously monitor and manage all of your business transactions, so that you can align IT operations with business goals and objectives. To understand whether an application is successful, you must understand the real-world performance of all business transactions as they traverse, the entire IT production environment. Only BTM provides this kind of real-time, cross-infrastructure visibility. This has a major impact on a number of core capabilities:
Complete auto-discovery: OpTier BTM can auto-discover each and every transaction type across every single tier without any manual development. Since the process is continuous, you always have an up to date view of the environment. OpTier also leverages this technology to calculate execution time between tiers which is crucial for detecting network problems. Since APM solutions have only partial discovery capabilities, if at all,it is necessary to manually update the model when there are changes to the environment.
Full data collection in production: OpTier BTM collects a full set of metrics even in high-transaction volume production environments, with less than 2% overhead on CPU. This enables early detection of performance issues and accurate problem identification and isolation. Deep dive solutions collect large amounts of detailed data that can be useful for problem resolution. But when they collect the full set of metrics in production environments, they can generate significant overhead. . As a result, they often require fault recreation in a non-production environment.
Problem resolution: OpTier BTM shows what happened on each tier even before the problem started and puts the tier metrics into the context of the affected business transactions. This enables you to isolate and rectify the issue without the need to recreate it. An approach that is made up of end-user experience monitoring and application component monitoring does not provide visibility into each step along the transaction path. In addition, deep dive solutions are employed after a problem occurs, and use sampling in order to reduce overhead. So while they can provide important insight into the cause of a problem on a specific tier, they will provide only a partial picture of what has occurred.
Problem identification: OpTier BTM detects application tier performance breaches before they affect users. Since OpTier BTM monitors all transactions and SLAs in real-time on each tier, organizations can proactively avoid outages and slowdowns. APM solutions depend on end-user monitoring to detect problems at the point when they are already affecting users. While deep dive solutions can alert on threshold breaches of individual metrics (e.g. heap size, concurrency), a breach does not necessarily indicate impact on a business transaction. This can cause a constant “red-light” situation where alerts are so numerous that they are ignored.
All transactions, no sampling: OpTier monitors all transactions without metric sampling. In contrast, APM solutions often use a sampling approach (e.g. metrics are averaged over a 15 second period). Potential problems can be missed due to the fact that they get “averaged-out” (e.g. quick spikes), which could lead to outages or degradation in service performance. Deep dive solutions capture large amounts of data that can be valuable for diagnostics – however, due to the large overhead, they will begin collecting all relevant data only after an alarm is triggered.
Visibility without instrumentation: OpTier BTM has unique “observed tier” capabilities for monitoring performance and collecting metrics on adjacent tiers - even when an agent is not installed directly on those tiers. OpTier BTM also provides tier insight – without agents – by analyzing log files from individual tiers.
Visibility into environmental issues: OpTier BTM detects issues in the application environment such as a misconfigured load-balancer, incorrect database routing, and performance issues with 3rd-party service providers. A user experience monitoring solution will not be able to highlight any of these issues since it monitors only the periphery of the environment. Deep dive tools focus on resolving underlying application issues, such as application code faults and execution environment (i.e. JVM/CLR) configuration issues, but do not detect issues beyond these tier metrics. APM does not produce a comprehensive dynamic map of the path of business transactions across multiple protocols with the ability to alert to changes in the topology of a business transaction.
Change management: OpTier provides out-of-the-box views and reports that enable organizations to understand scalability issues on every tier and to compare the impact of application changes on utilization, performance and SLA compliance. OpTier enables organizations to assess, in advance, the business impact of any change and after the fact it correlates service outages and business impact with recent change requests. Because APM only monitors the user experience and obtains deep dive data for exceptions, it does not have the information required to assess the impact of change on the performance and behavior of transactions in the new environment. While deep dive metrics from two different time periods can be compared, this data will be available from a different interface and will not have business process context.
Visibility into transactions not initiated by end-users: Problems often occur due to activities that are not initiated by end-users (e.g. DBA activity, system maintenance processes). OpTier BTM tracks all transactions, whether they originate from end-users or not. End-user experience monitoring solutions are limited to the flow of transactions that enter the environment from the end-user or via Web Service calls. Where non-user or Web Service initiated transactions are monitored, manual configuration is required, making the monitoring solution inflexible and requiring reporting and monitoring to be performed through the Deep Dive interface rather than the EUM interface.
Unified solution: OpTier BTM features a single management server, console and data store for an integrated, unified approach to managing transactions from the end-user to the back of the data center. If a vendor builds stovepipe integration between its EUM and deep dive products, the customer is left with two products that may not work together seamlessly, including missing or conflicting data. In many cases, vendors offer megasuites of 5 products or more in order to provide the same functionality as OpTier BTM. These megasuites are challenging and expensive to use, support and maintain.
