New Relic: Pros and Cons from real experience
November 08, 2021
New Relic: Pros and Cons from real experience
Score 8 out of 10
Vetted Review
Verified User
Overall Satisfaction with New Relic
New Relic is used across engineering departments to monitor how systems operate in both non-prod and production environments. For the component that I own, I use it to monitor Tomcat health (like the number of thread pools) and monitor the latency/error rate of the requests that my service handles. This helps us to understand how our system behaves, how much our system can handle, and help in capacity planning. I also set up important alert signals based on New Relic data and route the signal to Pager Duty that we have engineers on a routine schedule to pick up those alerts when something unexpected happened in production. This helps to quickly mitigate production issues that can surface to end-users
- Many important signals(for my case, signals related to the Tomcat server) come out of the box.
- Dashboard is easy to set up and share among multiple department.
- New Relic Query Language is simple and intuitive to use.
- Dashboard doesn't provide parameterization capability. So if you need to create the same dashboard to capture metrics of different environments/ or hosts, you need to keep a copy and pasting the same dashboard again and again.
- New Relic Query Language doesn't really support more complicated join operation between 2 (or more) types of signals (or it does support that, but in a convolute way). So, it's a hassle when you need to capture a more complicated signal.
- If you span the timeframe in the dashboard to be too wide apart, the dashboard may be slow to load or even crash the browser.
- New Relic help lessen the time we need to set up our monitoring system.
- It helps us to detect issue sooner and mitigate customer's complaint.
- It help us to understand system load and anticipate/plan a change in resource.
We route the signal that is captured by New Relic in real-time. We route it to Pager Duty alerting system so that in case there is a production incident, the engineer who is scheduled to pick up those signals can act on that prior to, and prevent the issue from being escalated to the end-users. We also use it to monitor how the system behaves after we deploy a new change to the production, in real-time
Instrumentation is simple for New Relic. Many important metrics come out of the box so our engineers spend minimum effort in doing so. This means it free us the resources needed so we can allocate more effort to pursue other projects. We also monitor the metric that New Relic captures when we migrate from on-prem to the cloud environment or migrate the system between different hosts. This helps us to identify any anomaly after migration quickly
Our DevOps team uses New Relic to capture important server health, like CPU usage, the number of threads/etc. Nearly all of these signals come out of the box so they spend minimum effort in doing so. They also set up their alert system based on these signals, which help them to detect any anomaly in servers in production, in real time.
Do you think New Relic delivers good value for the price?
Not sure
Are you happy with New Relic's feature set?
Yes
Did New Relic live up to sales and marketing promises?
I wasn't involved with the selection/purchase process
Did implementation of New Relic go as expected?
Yes
Would you buy New Relic again?
Yes