I know this is a meme, but just in case someone doesn’t actually know. CI saves literally thousands upon thousands of dev hours a year, even for small teams.
As annoying as it is when someone else breaks the CI pipeline on me, it is utterly invaluable for keeping the vast majority of commits from being able to break other people (and from you breaking others). I can’t imagine not having some form of CI to preventing merging bad code.
You should have seen my last job.
Hah, or my current one. Before we had CI you just directly committed to master (on SVN). It was incredible how unstable our build was. It broke basically everyday. Then one of the senior back end guys got promoted to architect and revamped the whole thing. Probably saved the company tens of millions dollars in man hours, at the very least.
Even better is when you restrict merges to trunk/main/master/develop (or whatever you call it) to only happen from the CI bot *after all tests (including builds for all supported platforms) pass. Nobody else breaks the CI pipiline, because breaking changes just don’t merge. The CI pipeline can test itself!
I often wonder if there isn’t some goodharty kind of local-maximum trap hiding in this…
And a lot of users’ frustration, especially on more niche platforms (Linux, ARM, etc.) - things look much better on release when the code have been regularly compiled and, hopefully tested, on all platforms, not just the one the lead developer uses.
Why waste time with CI when you can save on thousands of dev hours by limiting yourself to only one giant fuck off release every year!
/Taps forehead so hard it causes brain damage
It wouldn’t surprise me if this meme was made by an ops guy.
Ops loves CI systems, if the artifact doesn’t come from Jenkins (or friends) it simply doesn’t exist to us.
I’m also ops and I get it, it just seems like they’re shitposting.
I sure hope so ;) else I’m on the wrong /c
We only have serious IT discussions here.
Probably also causes lots of hours of maintenance and troubleshooting…but it’s a net gain in the end.
I can’t even imagine not having a ci pipeline anymore. Having more than a single production architecture target complete with test sets, Security audits, linters, multiple languages, multiple hour builds per platform… hundreds to thousands of developers… It’s just not possible to even try to make software at scale without it.
Multiple hour builds dear god 😵💫
If you fuck up the setup and deploy to multiple environments at once with each one set to rebuild an image/program things can get long. You really have to fuck it though.
“Leeroy Jenkins” is what my backend guys say right before they huck a major DB upgrade into prod without testing it in staging.
Always Friday at 16:59 right?
Right before a long weekend where Monday is a government holiday.
Also, Leeroy tried to optimize his PTO and hooked a backpacking trip onto the long weekend. He will be out all week and will have no phone reception.
But he will have chicken.
Our old Jenkins box is called Leroy, and my old place it was called Jankins. Thankfully we’ve moved on from that trash.
Real talk- I agree with this meme as truth.
The more and more I use CICD tools, the more I see value in scripting out my deployment with shell scripts and Dockerfiles that can be run anywhere, to include within a CICD tool.
This way, the CICD tool is merely a launch point for the aforementioned deployment scripts, and its only other responsibility is injecting deployment tokens and credentials into the scripts as necessary.
Anyone else in the same boat as me?
I’d be curious to hear about projects where my approach would not work, if anyone is willing to share!
Edit: In no way does my approach to deployment reduce my appreciation for the efforts required to make a CICD pipeline happen. I’m just saying that in my experience, I don’t find most CICD platforms’ features to be necessary.
This is pretty much what we do as well
All the build logic is coded in python scripts, the jenkins file only defines the stage (with branch restrictions) and calls the respective script function.
This means it works on all machines and if we need to move away from jenkins integration with a new ci platform would require minimal effort.
You’re not advocating against CI like the meme seems to be, but rather for CI builds to be runnable on human’s machines and the results should be same/similar as in when running w/in the CI system. Which is what CI folks want anyway.
What about related tools such as viewing artifacts such as for example total memory usage, and graphing that in the browser.
And sending emails, messages etc in case of a failure or change.
Most of those things mentioned aren’t bona fide needs for me. Once a developer is deploying their project, they’re watching it go through the pipeline so they can quickly respond to issues and validate that everything in production looks good before they switch contexts to something else.
I see what you’re saying though, depending on what exactly is being deployed, the policies of your organization, and maybe expectations that developers are working in another context once they kick off a deployment, it could be necessary to have alerting like that. In that case it may be wise to flex some features of your CICD platform (or build a more robust script for deployment that can handle error alerting, which may or may not be worth it).
I come from game dev. We do lots of checks on the data that all kinds of people can screw up. So it’s important these situations are handled automatically with an email to the responsible person. A simple change can break the game, or someone might commit an uncompressed texture so the memory usage jumps up.
Yeah, except for the Docker part
What’s wrong with Docker?
TBF, the problem isn’t Docker, it’s overused containerization
I’ve found Docker helpful when I want to use it to build binaries or use CLI tools that may not be available directly on the CICD platform. Also, Docker makes it easier to run the same code on MacOS that I ended up running on a Linux CICD server.
What would you consider to be overuse of containers?
Honestly, CI is only meaningful on bigger projects (more than 100 man-hours invested in total). So I most often go without.
But I do see its point.
I don’t think there is a single right or wrong answer but to play devils advocate making your CI tooling lightweight orchestration for your scripts that do the majority of the work means you lose the advantages of being able to easily add in third party tools that you want to integrate with your pipeline (quality, security, testing, reporting, auditing, artefact management, alerting, etc). It becomes more complex the more pipelines you are creating while maintaining a consistent set of tooling integrations.
Then you would probably enjoy concourse
Counterpoint: watching little green checkmarks appear when my PR passes a pipeline step gives me dopamine
I’m a bit confused. I thought “build system” referred to systems like autotools, scons or cmake. How are they related to green checkmarks? Couldn’t one also get green checkmarks when using a build shell script or makefile?
Is a CI/CD pipeline not a build system?
(this isn’t a “gotcha”, I genuinely may have misunderstood the post)
Ah, good 'ol Jenkins. It’s on my list of software I never want to use again, twice.
One feature was really sweet though: being able to edit the Jenkinsfile script inline and run it. On the other hand, that encouraged the wild cowboy lands. Contrasted to GitHub Actions, you get to see how many commits it took to get right 🙃
Nobody will see me force push to “bugfix/gitlabCI” the 10th time today…
What’s wrong with Jenkins? Works pretty great for automated scripts that need to run on a schedule, but I imagine you and this post specifically mean in reference to CI/CD
I work for a very large company which uses Jenkins for CI/CD and it’s an absolute nightmare. Granted, some of these issues may be related to how my company has it setup. I’m not in DevOps so I wouldn’t know. But these are my complaints:
-
Can have incredibly long queue times in some cases. It takes forever to spin up additional build agents to meet demand. In one case we actually had to abort a deploy because Jenkins wasn’t spinning up more build agents, and our queue times were going to put us outside of our 3 HOUR maintenance window.
-
Non-standard format for pipeline configuration files. It could just be JSON or YAML, but noooo, I have to learn something completely different that won’t transfer to other products.
-
Dated and overly complicated UI with multiple UX issues. I can view the logs in a modal from the build page, but I can’t copy from them? Fuck off Jenkins.
I’m actively pushing my team to transition to GitHub actions, because it’s just better in every single way.
Ah man, yeah I use it for a much more constrained and very narrow use case. We only use GitHub actions for CI/CD, it can be clunky itself in some aspects but otherwise works great.
And if you have a large company and many teams, you think actions will help? (Aside from the UI issues you mention). Rebuilding the Jenkins from scratch now would probably get rid of most of your problems, but in a year is gonna be a mess. It’s similar to how it’s going to go with and CI.
Also, a good DevOps person or team will keep the Devs happy (or at least, not very unhappy) with any tool, a bad one will suck anyhow.
At least that’s my experience.
The poorly documented pipeline scripting was always a nightmare for me, plus there’s two different types (declarative vs scripted) and so you have to be extra careful pulling examples from the Internet.
The build agent issue is 100% on your company not providing enough agents though. These days you can spin up agents as containers on k8s as needed.
-
Allow me to blow your mind: my Jenkins build calls build.sh because I’m not a fuckin idiot
WHAT THE FUCK IS
BUILD.SH, ALL MY CODE IS INMAIN.CJUST COMPILE THE DAMN THINGJoke’s on you. I have a Jenkins hook from github to trigger build.bat! :P
If you want to take Cargo away from me, you’ll have to pry it from my cold, dead claws. 🦀
I don’t think cargo is the problem. it’s idiomatic and it’s like “build.sh”
Cargo fetches dependencies, runs a variety of build tasks, can build a typical Rust project with little or no build scripting, and is configured with a straightforward TOML file. It’s not at all like a hand-written shell script. It’s also much more pleasant to use than any other build system I’ve seen, including shell scripts.
yea, as I said, it’s idiomatic. it replaces the need for a build.sh.
Is that not true of all build systems?
Yeah sure. Try building anything more complex than helloworld.c with a build.sh
The point is that “build.sh” implies a single file, which becomes an absolute nightmare to maintain on larger projects
I believe he goes by Leroy
Ha ha. I work on Bazel (a great build tools, https://bazel.build), and I agree 100%.
God I hate bazel/blaze.
Thanks? I’m not sure why you wanted to share that with me.
You shared randomly, they shared randomly. Balance in all things.
Hey buddy can you step over here, there’s a very tall cliff I want you to see
Please ignore everyone else being unkind - I’m somewhat new to build systems in general, what are the advantages/disadvantages of Bazel compared to other build systems?
If I break our master build in CI, I get multiple emails and people saying “fix this”!!! I wouldn’t have to fix it if you stopped letting people commit directly to master and stopped using git rebase! 😁
The build system issue is getting out of control. Just look at cmake
When your build system is a build system for build systems you know something went wrong years ago
I’ve been using Gearset for Salesforce CI/CD for a while and it’s pretty simple to get up and running and it just kind of works. I’m looking into integrating it with Azure for our .net stack but not sure how smoothly that will go.
