End to End Transaction Monitoring

The term "end to end" - in relation to transaction monitoring is very over used. "End to end" Website monitoring, "end to end" Server performance monitoring, "End to end" Database monitoring, monitoring traffic, "end to end" Network performance monitoring; the list goes on of tools that claim to provide "end to end" monitoring solutions when in fact they only provide monitoring for one small part of the bigger picture - the diagram below explains it all.

Why True End to End Transaction Monitoring is Different than Monitoring Server Performance

Monitoring server performance traditionally means that you are trying to make sure that CPU utilization and memory consumption are not maxed out, ensuring that resources are not maxing out does not promise true availability of applications. Traditional "end to end" monitoring tools provide a dashboard that shows the health of each one of the individual components within the data center. But what if someone forgot to re-connect a network cable after maintenance or what if a load balancer was configured incorrectly and is not executing a proper round robin? An alert will indicate that something is wrong with the servers that are still connected when in fact the problem is that traffic is simply not being equally distributed. You could implement some sort of network performance monitoring solution which would aid by monitoring the usage of the network, but that is not a single end to end solution anymore.

What if a runaway process in one server is "bombing" second server with requests? The server monitoring solution will blame the poor server that is being bombed by the server which has a runaway process running on it as opposed to indicating the source of the problem.

Real User Monitoring – An End to End Solution?

If you only take the user's tier into account, real user monitoring could be considered an end to end solution for that tier alone. You could also say that with real user monitoring you can see both ends of a transaction – hence the claim to provide "end to end". What about everything in the middle? What action is to be taken when a problem is detected by the end user experience tool? What if the root of the problem resides within the database or the mainframe? Shouldn't an end to end solution be able to take care of everything, including triage within the datacenter? Website performance monitoring is important for understanding the quality of service that your customers are receiving and there is a lot of value in that, but an end to end solution should also help you find the cause of the problem along with simply reporting its existence.

End to End Transaction Monitoring Tools Deliver

What do the following components have in common – the User Desktop, a Firewall, a Proxy, a Load Balancer, a Web Server, an Application Server, a message broker, a Database and a Mainframe? The answer is - you guessed it - transactions. The only way to provide a true end to end solution is by monitoring every transaction from the moment any user clicks any button and all the way through all of the different tiers and only business transaction management solutions can promise that.

Performance monitoring tools that are not showing end user performance are not focusing on what is most important to the business. Website monitoring tools that send synthetic transactions to the website and check response times are not showing what users are really experiencing or monitoring the usage of different services. Database monitoring is important but without knowing the context of that problematic SQL server transaction, resolution can be a shot in the dark.

Transaction Monitoring With Business Transaction Management

End to end solutions must be able to monitor all of the infrastructure's components. Transaction monitoring solutions do that automatically since they monitor the object that ties all of those different components together – the transactions. Business transaction management solutions enable a drill down that begins from each transaction type that is running on the system and ends with the smallest event that composes a single transaction instance. In this manner, not only are you ensuring that everything is running smoothly, but when things start to go wrong you can perform immediate triage and resolution.