Customers often ask me the difference between their switch log analyser (usually either homegrown software, Prognosis, or ESQ) and a transaction monitoring solution. It’s a fair question, since both are marketed as transaction monitors. Here’s the difference.
Switch log analysers only see transactions as they enter and leave a single component – your payment switch. They provide a deep, transactional view of the performance of a single component in the transaction execution path.
Transaction monitoring software sees transactions as they enter and leave multiple components. It provides a transactional view of the performance of your whole processing environment, including 3rd party participants. If you’re not using a transaction monitor alongside your switch log analyser, here’s what you’re missing:
Visibility into attempted transactions
Transaction monitoring software sees devices attempting transactions that may never be processed by your switch (and hence available in switch logs) – perhaps the message format is wrong, the device is not recognised as part of your network, or it’s having network connectivity problems.
If you’re driving a large fleet of devices, you know that the number of transactions you process is always lower than the number attempted. The goal is to narrow this gap as much as possible, because you’re paying, either directly, or indirectly for every one of these attempts, regardless of whether they’re successfully processed.
A shared problem isolation environment
What if the problem isn’t in your switch? Transaction monitoring software provides a complete view of the transaction path, including its transmission across the network (be it wireless, TCP/IP or dial-up activity over telco lines), it’s execution within 3rd party environments, and the overall outcome of the transaction. This allows multiple teams collaborating on an issue to work around a common view of the problem.
Field technicians can see details on how the devices are connecting, network engineers can see network timing information and TCP/IP details, application support personnel can see application timing and response codes, and line of business executives can prove to 3rd party partners that their systems received the transaction, but failed to handle it properly. And everyone sees this information in the context of an end-to-end transaction.
A faster way to monitor new services
When you bring up a new service, there’s often a difficult decision as to whether it’s “worth the effort” of extending your existing monitoring system to cover it. You have to roll out new agents, acquire new licenses for the servers, etc.
A lightweight transaction monitoring solution allows you to quickly include new services for monitoring without requiring extensive staffing resources, instrumentation, performance testing, and prohibitive licensing costs. Once you’ve proven the business case for the new service, instrument it with a switch log analyser to get deep performance information.
A single view across multiple switches
Are you dealing with more than one transaction switch? As a result of mergers and acquisitions, most payment processors have multiple switching platforms. The challenge is that they must appear to the front end customer like a fully functioning single integrated switch, but often it is chaos behind the scenes. Customer service and operations teams are logging into multiple systems to track down information, you get two copies of the same report for different systems, or worse yet, you get great visibility into one system, and limited (or no) visibility into the other.
Transaction monitoring software can provide a single, consolidated view of transaction performance across multiple switches and switching platforms. It can also provide centralised inquiry, troubleshooting, reporting, and alerting capabilities.
So are we saying ditch your switch log analyser? No. But recognise that it’s not a complete solution. It’s great for deeply analysing performance problems related to your switch. A transaction monitoring solution is great for understanding performance problems across your whole processing environment. Use the two together for best results.