Your data. Anywhere you go.

New Relic for iOS or Android


Download on the App Store    Android App on Google play


New Relic Insights App for iOS


Download on the App Store


Learn more

Close icon

Cloud Adoption with New Relic: Online Guide

rfb

#1

Cloud Adoption :cloud:

The earlier you instrument the faster you go.

Adopting the cloud is not simple or easy, but it also doesn’t have to be difficult. We see many customers that take the leap, only to find out that they made costly mistakes or negatively impacted the performance of their apps. Our Cloud Adoption framework is designed to help you avoid that fate. Instrument before, during, and after our migration to the cloud to ensure a smooth and successful transition.

In this online guide, we want to share with you how to leverage New Relic throughout your cloud adoption process. This is a new area for our community, so we’ll be building out content over time, and relying on you to share the experiences, tips, and ideas that will make this framework most valuable for others. We cant

So how do you do it? We’ve got you covered with these three phases: Plan, Migrate, Run.

Plan

Prepare for the move by setting performance baselines, Identifying dependencies, and prioritizing your migration order:

Step 1: Application Baselining
Establish a baseline and validate your existing performance.

Step 2: Identify Application Dependencies
Understand the impact of your application migration to the cloud.

Step 3: Prioritize Migration Order
Verify and ensure that your application is ready for the cloud.

Migrate

What you need to measure before, during, and after the move, so you can ensure success at every step.

Step 4: Identify Issues & Roadblocks
Detect errors or other possible issues.

Step 5: Validate Cloud Improvements
Validate that your cloud service use is successful.

Step 6: Acceptance Testing
Compare on-premise performance to cloud performance.

Run

Maintain and optimize your cloud apps and infrastructure by instrumenting everything.

Step 7: Monitor Cloud Services
Monitor, query, and alert on application and use of cloud services

Step 8: Cloud Spend Optimization
Understand and optimize cloud spend.

Step 9: Application Refactoring
Re-architect or refactor a cloud application.

Step 10: Optimize Customer Experience
Improve end-user experience for cloud application.


Cloud Adoption Solution: Migrate and Validate Cloud Improvements
Cloud Adoption Solution: Migrate by Identifying Issues and Roadblocks
#2

Welcome to our month of Cloud Adoption Explorers! We want to make sure that your journey is more :partly_sunny:️ than :cloud_with_rain:, so we’re asking you to share your experiences with the rest of our New Relic community. Your mission, weigh in on this poll and reply with an answer to the discussion question:

Where are you in your cloud adoption journey?

  • Born in the Cloud
  • We’ve reached some elevation, but we have a ways to go
  • We look up a lot and think about clouds
  • Feet firmly planted on the ground

0 voters

Discussion Question:

If you’ve been making the move to the cloud, what’s the BEST and the WORST choice you made so far?

Happy chatting!


#3

Don’t treat the cloud as just another data centre. Look at the features that are available such as elastic computing and auto-scaling as part of your application design.


#4

Excellent advice, @stefan_garnham ! :blush:


#5

Let’s talk planning.
We want to focus in on application baselining which is one way you can use New Relic to help plan a successful cloud migration. Let us know in the poll below where you are with baselining AND weigh in on our discussion question of the week below.

Do you Baseline, tho?

  • Yes totally!
  • No, but this community has me convinced I should.
  • I would rather not admit publicly…

0 voters

We wanna know:

What is your key metric? (You get double points for sharing any NRQL here!)


#6

Response time of transactions at the back end as these are easiest indicator of performance degradation. Can be average duration or apdex depending on the application responses :wink:


#7

100%, @stefan_garnham!

I am curious: do you have any Alerts set up to notify you and your team when response times slow way down?


#8

For day to day running we use Apdex alerts as this can provides us an indication of general network issues. We have critical transactions which have a variety of alerts on them for response codes, response times and specific exceptions where we identify downstream issues and reduce revenue loss.


#9

@stefan_garnham Please share those alerts with me please. Need to see what you got.


#10

I cannot give you the alerts but here is some of the NRQL used in alerts. Consider that we run mostly on Microsoft with .Net code base.

WCF service faulting:
SELECT count(*) FROM Transaction WHERE errorMessage LIKE '%System.ServiceModel.Channels.ServiceChannel%'

Check the percentage of downstream calls returning server unavailable. Has helped identify 3rd party API issues.
SELECT percentage(count(*), where errorMessage like '%The remote server returned an error: (5%' ) FROM Transaction

Similar but using a Gateway error check
SELECT count(*) from Transaction WHERE errorMessage like '%Bad Gateway%'

For Browser Ajax checks you need to use the metrics instead of a NRQL query. Ajax/StatusCode/500 is one we use along with 504

We are now looking at using NoticeError to identify particular events which require immediate attention. We then use a NRQL alert to look for those specific exceptions. It is an implementation we have only tried once but it has been effective.


#11

When you first got your apps in the cloud, was their performance:

  • Better
  • Worse
  • The same

0 voters

We wanna know:

If your initial move was less performant, what was the culprit? How did you find it?


Cloud Adoption Solution: Migrate with Acceptance Testing
Welcome New Explorers | May 21, 2018
#12

Bad software configuration caused bottlenecks where there shouldn’t have been any, just really simple stuff.


#13

Thank you @stefan_garnham this is a great share, I did not explore this realm yet and I have a starting point now.

thank you.


#14

Hi folks! Cloud month is rolling by, and even though it seems that ya’ll are more comfortable thinking about cloud adoption than talking about it (no tips? I KNOW YOU HAVE TIPS!), the poll results are pretty interesting so far:

Half of you are cloud native! That definitely surprised me, but it’s also great to hear. If we’re helping folks born in the cloud, we can help you get there too, and that’s what the rest of you are focused on!

Gold :star2: for the majority of you, but a significant chunk of you are still sorting this out, and that’s ok! In our work with customers, we found that many of them did not create a baseline before a move to the cloud, which is exactly why it’s one of the steps we wanted to highlight this month. Hopefully you all find the guide for baselines useful and put them into practice!

I love this week’s poll because it shows exactly what I would expect - a pretty mixed bag of results! Half of you saw better performance (:clap: well played!) but the other half saw no performance improvement or a decline in performance. That’s another thing we’ve witnessed across the many customers we’ve helped make this move. Without careful attention in the early phases, and without good baselines, it’s easy to see a decline in performance when you make the move. Acceptance testing is the only way to know!

Next week we’re going to talk about cloud spend, and I look forward to seeing more poll results and (ahem) conversation!


#15

How shocked were you when you got your first cloud bill?!

  • :sunglasses:
  • :hushed:
  • :sob:

0 voters

We wanna know:

Your best tip for managing your cloud expense now! Share your experience with us below.


#16

Luckily, I don’t see the bill as it is paid by someone else :sunglasses:


#17