Fork me on GitHub
#off-topic
<
2019-12-21
>
seancorfield00:12:16

@jayzawrotny New Relic is what we use. We love it!

seancorfield00:12:11

(specifically New Relic Infrastructure -- runs a small agent on each server, and provides detailed dashboards via a web UI of all sorts of stuff, plus configurable alarms)

jaide00:12:55

Thanks, it seems like a strong candidate!

seancorfield00:12:01

We use it on all our Clojure web apps as well, both frontend and backend. It can produce a huge amount of analytics and automated monitoring/alerting. Happy to answer any questions you might have.

jaide00:12:46

The long-short of it is that we're migrating our Django monolith on a local data server -> cloud hosting -> Clojure. However, we need to know how much it's going to cost to host. To do that we need to know what our main CPU, Memory, and Disk resources are on average as well as the maximums. We only need it for about a month so we can make an informed decision. Do you think New Relic is well suited for that? We're considering trying https://www.netdata.cloud/ first.

seancorfield00:12:40

I'd certainly get a trial NR account and see what it gives you. Pretty sure they have a Django plugin too so you can get all sorts of insight directly into that app beyond the infrastructure stuff.

hiredman04:12:09

newrelic is really nice. I wish it gave me options for aggregates. As far as I have seen you get either a total or an average, and sometimes I'd like a 90th percentile

vemv12:12:13

> specifically New Relic Infrastructure nice stuff. I'm currently in the process of introducing this in production I wanted it most of all for gathering JMX metrics such as GC stuff. Hope it works nicely!

deleted04:12:56

I don't like that basically every metrics collection service does multiple levels of aggregations of values to the point that it's impossible to know what values were actually reported

adamfeldman15:12:54

https://www.honeycomb.io doesn’t pre-aggregate, and allows you to slice and dice across high-cardinality fields

kulminaator08:12:58

I didn't like datadog when they sent our their breach email

kulminaator08:12:17

If you expect people to install your agent applications in their servers you must not have breaches.

kulminaator08:12:22

@seancorfield has new relic revised it's pricing plans ? Last time i checked they charged way too much per node

4
kulminaator09:12:28

Last time i checked new relic were charging more for monitoring than aws was charging for running their micro or small instances. That just doesn't make sense on paper :-(

dominicm09:12:10

We just use aws cloudwatch

dominicm09:12:30

You can even just point it at an scaling group

sova-soars-the-sora15:12:31

Feature Request: heat maps for null

seancorfield19:12:47

@kulminaator I'm not sure what price we pay -- but we use a lot of their services across the entire suite and management gets a lot of value out of it.