Are there any better CI services than Jenkins for on-premises?

Hi everyone,

​

I want to migrate from Jenkins' CI pipeline to another service. I'm tired of using Jenkins for plugin dependencies, poor documentation, and deprecated plugins.

Our team of 10 people is using Github Enterprise and we want to run our CI pipeline on our machine.

​

Some requirements are as follows:

\-Commercial solutions are ok.

\-If it is open source, the quality of the document should be much better than that of Jenkins.

\- It should be able to run builds on our machine using Linux and macOS.

\- Polling (PR, Branch) must be available for the GHE Repository.

\- Pipeline as Code is required, but Configuration as Code is not strictly required.

\- Code has many dependencies (about 20GB), so local cache management should be possible.

​

Some of the candidates I've found are Github Action (not yet available on GHE), Cirrus CI, Azure Pipeline but I haven't looked into each candidate deeply. Are there any other alternatives?

https://redd.it/k4yihw
@r_devops

I'm a new employee at a new team that adds features to exisiting, changing code base, should we have build/deployment permissions?

So we're a newly hired 6 member team adding new features to a program that is also being handled by multiple teams. Problem is our team doesnt have permission to build for QA and waits for other teams to create a build so that ours changes get included. This is becoming a bottleneck since QA can't test features from our team immediately since deployment to QA is done manually. Should I talk to the higher ups to automate deployment upon PR for the teams? Will this create confusion which is the correct version for QA to test since at the end of the day it's all the same huge application? This is my first time working with a huge codebase that is being handled by multiple teams so I'm thinking of ways to make our integration faster. Will be really grateful for some advice. Thanks!

https://redd.it/k4ordu
@r_devops

What are best videos or blogs to learn GCP from scratch??

Gcp beginner, please help me

https://redd.it/k4lt7x
@r_devops

whatsupstream - a golang-based tool for oss devs/enthusiasts which runs in background and notifies you whenever an issue (with certain labels/creators of your choice) is raised in the repos you want to contribute to.

Here's the tool - [whatsupstream](https://github.com/yashvardhan-kukreja/whatsupstream)

So, say, you want to be notified whenever an issue with the label "good first issue" or/and "sig/node" is created by some user "alex" (or any other user) in the kubernetes repo, whatsupstream can notify u on ur desktop and u can pick it up before anyone does.

With it you dont have to keep on checking the repos and refreshing for the issues u wanna pick up.

---

So, I faced this problem while starting to contribute to kubernetes. I wanted to find "good first issues" but whenever I found one, someone already had picked it up and I couldn't just sit everyday clicking and refreshing the repo to find and pick up "good first issue". And I felt that this must be a pretty common issue.

So, as a tribute to the oss community, I coded this tool. [whatsupstream](https://github.com/yashvardhan-kukreja/whatsupstream). I even used it for a while and It worked like a charm and notified me whenever a good first issue came to kubernetes and I was able to pick them up and work on them.

Do check it out and give some stars too😂 if you like it. 😄😄

PS: Constructive criticism is always welcome :)

https://redd.it/k51p92
@r_devops

you may want to take some classes with Udemy Cyber week

Hey, guys, [Udemy's cyber week](https://www.tryblackfriday.com/2020/11/udemy-cyber-monday.html) is here, you may want to take some classes at lowest price of the year..

https://redd.it/k52xwy
@r_devops

I built a cron job monitoring tool after a scary incident

I hope this kind of post is allowed here.

I have a simple application which posts to social media sites like facebook on a schedule basis.

I slowly started realizing that all the posts are being posted twice. I initially blamed my code and spent hours figuring out the root cause but of no use.

After further investigation, I found out that the cron job which posts them to social media sites is running on two servers. This extra server is a dummy server which i started to migrate the application. But i left it behind in the middle to focus on it when i have more time. But i did not realize that the cron job is running on this server.

I immediately stopped the cron job from the dummy server. Then i thought,

* what if this cron job is processing payments?
* what if this cron job is generating shipment labels for e-commerce store?
* what if this cron job is doing a very critical task which should not be duplicated?

I thought others might be in the same boat and realized we should start monitoring our cron jobs.

​

So i built [CronBuzz](https://cronbuzz.com/?utm_source=reddit&utm_medium=devops&utm_campaign=post-12-01).

CronBuzz monitors your cron jobs and alert you when your cron jobs stopped running, duplicated, or running longer than they are intended.

Are you using any kind of cron job monitoring tool? Please let me know what you think of this tool.

https://redd.it/k4z1kn
@r_devops

Simplifying container management in Docker Swarm, Kubernetes, ACI and Edge environments

If you haven't yet discovered Portainer to simplify your container management, grab the free community edition (it's an open-source project) from [https://www.portainer.io/](https://www.portainer.io/) and let us know what you think.

If you're curious about container management within a business environment, catch the launch of Portainer Business on YouTube Premier. [https://www.youtube.com/watch?v=YiNyx\_qhUOQ](https://www.youtube.com/watch?v=YiNyx_qhUOQ)

https://redd.it/k52dr8
@r_devops

Dave's article: "Burn your ticketing system" - thoughts?

One of the best articles I've read in a while: "[Why you should burn your ticketing system](https://techbeacon.com/enterprise-it/why-you-should-burn-your-ticketing-system)" by Dave Mangot (not associated to him or advertising here, really just like the article), he sent it over this very morning. In the article he argues that there are three essential reasons why you should not have a ticketing system for ops (the kind where you would request another instance of a Postgres DB etc):
\- ticketing systems require waiting, waiting is waste
\- tickets in the system should be viewed as software exceptions being thrown by the value stream
\- Toil! You'll end up hiring a huge amount of frustrated, overqualified people doing boring, repetitive tasks.


Thoughts on this? I'm seeing similar situations a lot at different companies. I'm always wondering: how do you determine what the right level of automation is? From what level of repetition onwards should you spend how much time? Would love to hear your thoughts!

https://redd.it/k552mz
@r_devops

How to implement performance testing in CI pipeline? or just how to implement performance testing in general?

Hi! Lately we've seen huge downtime because of performance issues. One of the developers merged really bad code into production that took entire DB down for hours.

Today I've been wondering what if there is an actual tool that does performance testing at the CI level, that would be awesome. Like similar to how we do unit testing, there could be certain threshold that needs to be passed for that test.

Or if it's too fictional, where should I start to prevent such issues.

Thank you.

https://redd.it/k4zra1
@r_devops

NEW RELIC - HOW TO ADD AN EXISTING SYNTHETICS ping MONITOR TO A LOCAL New Relic APP?

So I'm relatively new to New Relic.

I have a synthetics monitor i created for a website, a simple ping one.

I then went to create a new relic app locally, which is new to me as well. How do I incorporate the above synthetics monitor I created successfully, into my new relic app?

I'm going through a documentation and videos and it isn't clear to me just yet.

I also want a very basic new relic app created right now, a space for an application to observe things going on with the site, so I'm trying to do the bare minimum setup for right now.

If I understand correctly, a new relic app is created locally in a local instance of new relic, and then when I publish/deploy it, New Relic will host it on our actual account so that the new relic app is no longer only local to me? This deployed app will run with the code from the New Relic infrastructure, and won't need to have it started on my machine locally (`nr1 nerdpack:serve`) in order to run and be there for others to see in the new relic account?

New relic, the new relic app and the file structure for it is all new to me, as well as the monitor.

PS: apologies, I don't know why the title of the post would only do cap locks automatically, for most things, it apperas. odd I guess

https://redd.it/k4ude6
@r_devops

How to get over the initial pain

Java and PowerShell are F'ing hard things to get started in. Im slightly discouraged and that makes me angry.

I need a pat on the back or something. Please help.

https://redd.it/k4l77b
@r_devops

I am kinda new to this. How to push 2 or more github repositories to google cloud build and run

So I have full stack nrwl application which is sperated back end , front end and one microservice. All connected docker-compose but tried it locally. I dont find a example of being run at the same time one to another. Any suggestion appricieted.

https://redd.it/k4kepy
@r_devops

Introducing Project Shipwright (Part 2) - Build Container Images on Kubernetes

About three weeks ago I wrote a [blog post](https://shipwright.io/blog/2020/10/21/introducing-shipwright-part-1/) introducing Project Shipwright, and posted the link [here](https://www.reddit.com/r/devops/comments/jrkwbs/introducing_shipwright_a_framework_for_building/). The first post set the scene, explaining why we started the project.

Yesterday I posted [Part 2](https://shipwright.io/blog/2020/11/30/introducing-shipwright-part-2/) of my series, which introduces the Build APIs and explains how they work together. Please take a look, and let me know what you think!

https://redd.it/k4kbve
@r_devops

Introducing microservices and Docker with a whiteboard and a marker pen :)

Hi team! some months ago I made a very educative and visual video explaining **microservices** and **Docker** from the perspective of being interesting, fun and useful. I just want to share the content here. I hope some of you may found it interesting and useful. Thank you and have a nice day!

[https://youtu.be/D3okLNBL1lE](https://youtu.be/D3okLNBL1lE)

https://redd.it/k4krsy
@r_devops

What is Jenkins?

Let's start with an introduction to one of the most popular [DevOps tools](https://www.janbasktraining.com/blog/devops-tools/) that are popular as Jenkins. [Jenkins wiki](https://wiki.jenkins.io/#all-updates) is an open-source tool that has many plug-ins and it is written in Java. The Jenkins wiki tool was launched to build and test software projects in an easy way.

Developers can easily integrate application changes with this tool to help the user to obtain a fresh build. The software can be tested and delivered continuously with the help of various integration and deployment technologies.

Through automation, software developers can accelerate the process of software development. Jenkins mainly integrates all the stages of the software development lifecycle that are documentation, packaging, testing, deployment, static analysis, and other ones.

[Jenkins](https://www.janbasktraining.com/blog/jenkins-interview-questions/) plugins help the developers in providing continuous integration and various stages can be integrated through Jenkins. To integrate any specific tool like Git, Amazon EC2, Manen 2 project, HTML publisher, etc. you can download the appropriate plugin and integrate the tool. It is a basically advantageous tool and the reason for the advantages of Jenkins is:

* Jenkins is an open-source tool that has wide community support
* Installation of Jenkins is quite easier
* A vast number of plugins are available; even if any plugin doesn’t exist then you can develop it and add it to the community.
* Due to Java's written code, it is portable to major platforms.

https://redd.it/k4je5c
@r_devops

Puppet CIS Benchmarks

Hi all. I'm looking to push CIS benchmarks via Puppet to our infrastructure. I did see the CIS Compliance Service. Can anyone speak on it's success? Or, have you found success through a different route? Cheers!

https://redd.it/k5fmjg
@r_devops

Runbook Automation

Hi,

Have you considered automating runbooks? If so, which tools did you use to do so? I'm exploring Rundeck but some of their features like PagerDuty, Okta integrations come with a license fee of at least $20k.

Thank you.

https://redd.it/k5ilyw
@r_devops

Running vault in k8s vs dedicated vm

In my previous company we ran vault on dedicated hardware, so we had a few VMs in separate regions - per environment.

In my current company, I have an environment running one k8s cluster with a few nodes for services, but 2 nodes specifically dedicated to vault, running in separate regions, and am running 2 vault pods there.

Since my region spread is identical, are there any reliability gain from dedicated VMs over my current setup?

I have an urge to say dedicated VMs would be more reliable, but I don't have any evidence why. In fact hashicorp's own website provides a [k8s reference architecture](https://learn.hashicorp.com/tutorials/vault/kubernetes-reference-architecture) in addition to [standard reference architecture](https://learn.hashicorp.com/tutorials/vault/reference-architecture)

Also - could there be performance gains from VMs over pods in k8s?

https://redd.it/k5kowe
@r_devops

AWS - change file system on ec2 ebs root block device

Hello.

I've been looking for a way to change the file system in an EBS root block device with no luck. I'm looking to create a new EBS backed AMI with btrfs on the root block device. I'd prefer to do this as part of an Image Builder pipeline, but any method (except Packer) would help.

https://redd.it/k5quny
@r_devops

I'm looking for your feedback! (experiment about VR and TravisCI)

I am working on a very early idea: applying virtual reality in software engineering. My first try is this. I represent in VR the data that can be retrieved from [TravisCI.org](https://TravisCI.org) using its API v3. Your feedback will be very valuable to me. Thank you very much colleague

The experiment [is here](https://rv.dcc.uchile.cl/)

https://redd.it/k5ozve
@r_devops

debug a Docker container on aws Elastic beanstalk, still up but not working, no printing any logs or errors

Pretty much the title.

I have a simple python app doing some stuff in a main loop (while True: ...do stuff... sleep 5 sec). It runs in the official python 3.8 slim-buster image. I run the container on AWS Elastic Beanstalk (EC2 t3a.micro).

Now the app stopped working last week. I logged in the AWS console, nothing weird. I ssh'd directly into the EC2 instance, the container was still up with an uptime of 2 months. When I displayed the logs from the container, it indeed stopped logging last week. But what's weird is that it stopped in the middle of the algorithm (very basic stuff, nothing CPU/IO/time intensive), in the main loop, without logging any errors. It just stopped there. On the side of the EC2 instance, CPU usage is low, volume almost empty (I didn't think about checking the memory usage.. but if it was close to 100% I wouldn't be able to ssh there. edit: 50% so far but it has been up for only 30 min. Need to wait to see if it increases. edit2: doesn't seem to leak).

The same issue happened 2 months ago, so that is the second time in like 5 months that I've been running that app. If we exclude these 2 dates, the app did its job, 24 hours a day, 7 days a week.

**So I tend to think the problem is not in the code or in python otherwise there would be some errors and it would probably happen more often. So I guess it's Elastic Beanstalk or Docker.**

Any idea what it could be or how could I find the source of the issue? How to debug it?

Should I just force it to restart automatically like on a daily basis since it seems it happens after running for weeks? I don't mind the downtime, it's stateless and takes a few seconds to restart.

https://redd.it/k5k2cl
@r_devops