Monitoring

Just as the heart monitor in a hospital that is making sure that a patient’s heart is beating and at the right beat, software monitoring will watch computer metrics, record them, and emit an alert if something is unusual or that could make the computer not work properly happens.

You cannot fix or improve what you cannot measure?is a famous saying in the tech industry. In the age of the data-ism, monitoring how our software systems are doing is an important thing.

Web stack monitoring can be broken down into 2 categories:

• Application monitoring: getting data about your running software and making sure it is behaving as expected
• Server monitoring: getting data about your virtual or physical server and making sure they are not overloaded (could be CPU, memory, disk or network overload)

Here are few famous monitoring tools:

NewRelic

NewRelic requires you to add a piece of JavaScript to your website, this agent will collect information and send it back to the New Relic servers. It will give you a detailed analysis of how quickly your website loads in a browser, with a detailed analysis at every level of the stack. If the website is loading too slowly or users are experiencing error (500), there is a feature that allows you to trigger an alert. NewRelic now does much more than this, I’ll let you discover the rest.

DataDog

DataDog allows you to measure and monitor everything with graphs. It gathers performance data from all your application components. The service has a lot of integrations. You usually just need to properly configure your alert and you are good to go with solid monitoring coverage.

Uptime Robot

Uptime Robot is a simple service that will check that your website is responding from multiple locations in the world. This is the bare minimum when you host a website.

Nagios

Nagios is an open source project started in 1999, it is among the most widely used monitoring tools in the tech industry. It is, however, seen as outdated. Nagios had trouble adapting to the rise of the Cloud but is slowly trying to catch up.

WaveFront

Wavefront is a cutting edge monitoring service funded by great software engineers who’ve built monitoring tools for the best tech companies in Silicon Valley. The idea is to be able to analyze anything that can produce data points. A query language that looks like SQL allows users to apply mathematical operations to these data points to extract values or detect anomalies from the time series data. While it takes some time to get used to the tool, it’s the type of monitoring that the best companies are using. To my knowledge, LinkedIn, Facebook and DropBox are using a very similar tool for their monitoring needs.