Great for new applications
February 03, 2017
Great for new applications
Director of Development
Cage DataComputer Software, 1-10 employees
Score 7 out of 10
New Relic for Web Apps
Overall Satisfaction with New Relic APM
We use NewRelic on web and mobile applications to monitor performance of the applications. We also use their server monitoring tool to track system metrics for our clients systems to ensure that they are operating at peak performance. In the past, we would launch applications and not really know how they performed when users actually used the applications. NewRelic gives us near-realtime insight into what's happening and where the bottlenecks are performance wise.
- Integration is extremely easy. In our Ruby apps, we just include a Gem and a config file.
- Building alerts off of metrics is straightforward.
- NewRelic does a really good job of allowing users to look at different metrics for a timeframe to narrow down what effects crashes or performance spikes have on the applications
- Monitoring non-custom software with NewRelic isn't possible. For example, there's no way to monitor when a service like Nginx or PostgreSQL goes down
- The alerting system is split in two currently. Simply monitoring uptime of applications (is NewRelic receiving data?) is a separate system than alerting on metrics
- Crash reporting doesn't seem to pickup all crashes.
- I don't have to think about my monitoring. NewRelic is plug and play with our applications so it's a non-thought to get monitoring
- Integrates with the other tools we use. NewRelic has a huge ecosystem of tools that they integrate with so we don't have to pick tools based on what NewRelic works with.
- Time spent debugging issues is significantly reduced by easy access to metrics.
NewRelic is definitely built towards applications that you are developing. Datadog starts with an application installed on the server and then metrics can be fed into there. NewRelic also is much more plug and play than Datadog and Elk. If you want to monitor internal application metrics, NewRelic will just pick it up for most languages and frameworks. Datadog and Elk need to be built out for it. Datadog, however, is much better at monitoring and alerting on server based metrics or applications that can't be directly tooled with NewRelic APM.
If you are just getting started on an application, NewRelic is a great solution to drop in and start collecting metrics. The base metrics that NewRelic collects are suitable for most applications and the alerts can cover most scenarios. As far as alerting goes, NewRelic still does have a number of limitations. It's not possible to alert when a metric breaks out of the average for an extended period of time and if you have a lot of dependencies on external applications which can't be tooled by NewRelic, monitoring becomes difficult.