3 Takeaways from DevOps.com Better Together: End-to-End Software Delivery with Best-Of-Breed Tooling
A modern suite of best-of-breed tools is behind every successful piece of software. But, building a custom DevOps toolchain that fits your software, team, and organizational needs is easier said than done.
If you haven’t already, watch this DevOps.com webinar to see industry leaders from Atlassian, CircleCI, and New Relic come together and discuss how having the right integrations, architectures, and insights can protect you from challenges during software development and delivery.
We know you’re busy, so we compiled three main takeaways, the FAQs from the webinar, and bonus resources to get you up to speed. Enjoy!
If you have any additional questions following this webinar, feel free to submit them in the comments section.
On to the takeaways
In answering the question “But why DevOps at all?”, Camden Swita, Product Manager at New Relic, explains what DevOps really is about at its core. “On the surface, yes, DevOps is this continuous flow of information up and down the software lifecycle. This allows creators to get quick feedback so that they can design, plan, and build to maximize value for their customers,“ he states. However, he continues to explain that “DevOps has other kinds of deeper meanings, especially about working by certain values.” Camden reveals that those values are: 
🏘️ 
.
1. Unless you’re the one swinging it, don’t get too caught up on which hammer to choose.
No organization knows exactly what tools a team needs and so it’s best if developers pick their preferred tools from the open market. This enables more productivity and efficiency. While DevOps introduces a new way of working, it doesn’t necessarily mean that you need to implement completely new tools. Customizing your tools to what best suits your team’s current needs is a great starting point for implementing DevOps into your workflow.
Warren Marusiak, Open DevOps Technical Evangelist at Atlassian, demonstrates how Atlassian’s Open DevOps solution can help software teams focus on developing and building software. “The goal of Atlassian Open DevOps is to allow developers to set up integrated tool suites and bring in integrations with tools that they're either already using or they're really interested in using,” he relays. You can enjoy the benefits of a best-of-breed toolchain and the coordination of an all-in-one. Warren explains how Atlassian’s Open DevOps provides users with five tools to start with, including Jira Software, a software development tool that enables teams to effectively organize and deliver work through project management and issue tracking. “This is the central point of integration for everything. Your code tools push updates into Jira Software. As you make changes in your code that automatically populates your issues as your alarms fire, you can have them raise your issues for you automatically so that you can assign people to them and respond to it,” he states.
“When you have a successfully integrated toolchain, you will see improved developer metrics such as feature velocity, mean time to resolution, and improved developer work-life balance,” adds Camden Swita. The team later demonstrates several developer metric examples in the product integration walk-through.
2. Good fences make bad neighbors (and software) 🏘️
When it comes to software delivery, removing barriers and allowing teams to manage their entire software lifecycle improves the team’s performance, morale, and the quality of their product.
“It's critical to understand and have insight not only into production but also from creation to production. And that is where our integrations with world class technology partners like Atlassian and New Relic help simplify collaboration and deliver actionable monitoring insights when needed,” explains Rick Burta, Senior Software Engineer at Circle CI.
The ‘Better Together’ product demo shows how Atlassian Open DevOps, Circle CI, and New Relic products work together to enable you and your team with added visibility and automation in your workflow. Glenn Thomas, Principle Partner Engineer at New Relic, explains how the integrations were useful even in the process of creating the demo itself. “It actually proved out the value in having not only each of these sets of tools but also the integrations between them to troubleshoot,” he states.
3. Nobody wins a game of telephone
The feedback loop between development, design, and customers should be immediate and driven by observed facts, and developers should use data to make decisions and prioritize work. This will help your team ensure a quick and efficient delivery of high-value updates.
This is an important factor to keep in mind when deciding what tools will serve as the most efficient for your DevOps needs. When explaining this theme in the webinar, Camden Swita states, “your tools when taken together should be able to help you plan, build, integrate, deploy, observe, report, and repeat without a lot of nudging and manual copying and pasting in between those steps.” This is a great guiding principle to assess your current and future tools by.
FAQ
Do I have to use Bitbucket with Atlassian Open DevOps or can I use GitHub or GitLab?
Out-of-the-box, Atlassian Open DevOps provides Bitbucket and Bitbucket Pipelines for source control, pull requests, and CI/CD pipelines. But, all the features that you saw with updating the Jira issues in Bitbucket are fully supported with GitHub and GitLab as well. If you wanted to use either of those, by all means, integrate them and you can use their CI/CD products or you can integrate them with Circle CI or other tools as well. Atlassian Open DevOps wants teams to bring their tools to their pipelines and not be restricted in what they have to use. So, we're going to give you guys that start up set of tools that you can use and you can swap out any of them for what you prefer. You can also integrate additional capabilities as you have needs.
What would be a good process for building a DevOps pipeline or a good starting point to start with CI/CD?
Any development process that is static, tedious, and/or prone to human error is a great candidate for automation and putting into your CI/CD pipeline. Running tests qualifies as tedious, updating a release image qualifies as status, and deploying an application to marketplace or registry is definitely prone to human error at least for me. Those are the first places I'm gonna start.
Let's say I use only one of these one of these three tools. (Circle CI, New Relic, Atlassian Open DevOps). Describe today, how long would it take to set up this toolchain as we've demoed and how much would it cost?
New Relic:
From New Relic's perspective, it's relatively low cost. Because a free account at New Relic essentially gets a free hundred gigabytes of stores per month and we have a pretty streamlined data onboarding process, we've been able to show through the early stages of the Circle CI integration that logs just take a few minutes to set up and there's a lot of views that are prebuilt for users. And, combined with that a hundred free gigabytes per month, it means setting up New Relic with Circle CI remains free or nearly free type of experience for new users.
We barely touched the free tier. We were pushing it close to one gigabyte for all the work that we did throughout this process. One of the benefits is that you can start very small and assess value out of those initial integrations, connections, or sets of data just to see what does this do to our processes, and what does it do to our visibility? And we've given the free tiers for that very reason. We want you to experiment and not be constrained by cost, find out what works for you first, and then decide where to go from there. We have guides that'll help you as well with if you wanted to go bigger to look at the costing.
Circle CI:
On the Circle CI side, we did quite a few tests to set up this webinar and build the demo out. We just revised our free tier offerings and we reached about 10% of the free tier usage. So, at least for what we did today it's likely that it would be free for quite some time.
Atlassian Open DevOps:
We have a free tier for up to 10 users in Jira Software, Confluence, and OpsGenie and up to five users in Bitbucket. So, for small startups and new teams that are just getting the ball rolling, it's easy to spin up an Open DevOps instance and get proof of concept on, build some stuff out, and get it working in your organization for free. And, with that you can do all the integrations we've seen in this demo or other places as part of that, so the integrations are not restricted. If you have a small team, as a skunkworks project, to see whether it's gonna work for you or not, it's free to get started and free to try.
How does Bitbucket Pipelines work together with Circle CI?
Bitbucket will run whatever CI/CD pipelines there are configuration files for. If you're in Bitbucket and you've enabled Bitbucket pipelines and you have a Bitbucket pipeline YAML file, when a push hits the repository, Bitbucket will try to run that deployment. If there is a Circle CI integration installed and you have a Circle CI configuration file, when a commit hits that repository, it will try to run the Circle CI/CD configuration. So you could have either or both within the same project in the same repository. Theoretically, you could have both kickoff and do specific jobs. If there's something that you like doing in Circle CI that you don't like doing in Bitbucket pipelines, you could configure it that way if you wanted to and vice versa. It's very flexible. They inter-op nicely. Now, it'd probably be a terrible idea to have both of them doing the same steps. I wouldn't want to do two deployments to the same environment with two different CI/CD pipelines as that's probably not going to work very well. But, I think the tools would not limit you in that case. You're free to use them together.
What would you advise an org moving towards DevOps who need to take an incremental approach? What part of DevOps, would you advise doing first, then next?
Warren Marusiak, Senior Technical Evangelist for Open DevOps- Atlassian:
This is the goal of Atlassian Open DevOps, to give teams that really straightforward starting point. You can sign up for Atlassian Open DevOps, and you've got five tools out of the box that'll give you most of what you need for running DevOps. And, from there it's really easy to branch out. So, you could start with Atlassian Open DevOps, get your stuff up and running, and then as you have new needs on your team, integrate the products that fulfill those needs and you can do that incrementally one piece at a time and slowly kind of grow out your solution as you have capacity to onboard new tools. That's where I would start.
Glenn Thomas, Principal Partner Engineer- New Relic:
I would advise experimentation. You don't know what you don't know. And so, all three companies have given the ability to go in and experiment to turn things on, you can easily turn them back off again if it doesn't provide value. But, we want you to try to find value in what we're offering, so we give you the ability to do it very incrementally. I would definitely start with the DevOps tool sets, but very quickly you're going to want to see things. I would think that, with observability, organizations want to see their data come through and see what visualization can happen within each of the products. Both Circle CI and Atlassian have visualization components in them as well.
Rick Burta, Senior Software Engineer-Circle CI:
Address the parts of DevOps you consider doing first. The next, as everybody says, is we offer you an incremental way to do this, but if you're looking for specific steps, identify your pain points. What is the biggest problem in your development process? What hurts the most? What is the most unreliable or what leads to the most problems? My tried-and-true approach to doing this is I always start with my tests because I get very attached to having my tests run on pre-commit hooks. I make sure I have that in my CI pipeline, and get going from there. There are many ways. If your deployment process is really painful, knock that down next. To touch on Glenn's answer, experimentation is important. There are a bunch of times when you should experiment such as when your project grows in size and complexity, you learn new things about your system and the parameters that go into it. Or, when your processes change. Those are all great times to experiment, review what your new pain points are, and take this very iterative, incremental approach.
Who doesn't love more resources?
Here are some helpful links to help you get started.
-
Product: Atlassian Open DevOps | Try for free | Learn more
-
Product: Atlassian Marketplace | Discover apps and integrations for Jira
-
Product: Jira Software | Try for free | Learn more
-
Product: Bitbucket | Try for free | Learn more
-
Product: Confluence | Try for free | Learn more
-
Product: Atlassian Developer | Learn more
-
Product: CircleCI Integrations | Learn more
-
Product: CircleCI Integration with New Relic | Learn more
-
Demo: Atlassian+CircleCI+New Relic Product Demo | Better Together Demo
Was this helpful?
Thanks!
Tanvi Kaur
17 comments