To code review or not to code review?

I just got promoted to a team lead position (2 weeks ago). The 5 person team I joined has developers who have been at the company ranging from 10 to 20 years. I've only been at the company for \~5 years.

I think its critical that we have code reviews. I've always looked at what went to prod even before I was a team lead. I could easily enforceable code reviews via git requirements if I want to. I'm of the opinion that spending 5 minutes reviewing code will save 30 minutes in the future. Either by understanding the flow of the code more, finding/noticing bugs, or even just being able to standardize on coding practices.

However, should I enforce it? They feel that "our time is better spent elsewhere." And while the topics they bring up are considered things that we should spend time on. I think there is very little that is more important in a DevOps environment than code review.

https://redd.it/k9ms00
@r_devops

Effing Shell Scripts 2

For a while now, I've been working on a side project and thought it would be a good day to share it.


[https://github.com/madflojo/efs2](https://github.com/madflojo/efs2)


This tool tries to solve the area where you want to automate the installation or setup of something. But, it's not worth going through the hassle of setting up a complex config management system like Puppet, Chef, Salt, or whatever your poison is. The idea is you create a file similar to a Dockerfile in a directory/repo and run the efs2 command against that file supplying a target host. No primary server, no complex setup, just stupid shell commands and one executable on your laptop.

https://redd.it/k9nqez
@r_devops

A better way to build on AWS

So AWS gives startups $100k in free credits. Google and Azure have similar programs for startups. Then why isn’t every startup CTO starting on the Big Cloud?

The real cost of going with AWS is its complexity. There is hardly anything more important to an early-stage startup than moving fast, but this is exactly where AWS fails startup founders. It is hard to set up and manage, which is the opposite of fast.

This is why simpler tools like Firebase and Vercel are so popular. The speed of experimentation is unmatched. You can go live with a new product in a few clicks. But then you won’t use these tools at scale any bigger than a small experiment. It gets expensive very quickly, and workflows for teams aren’t adequately supported (perhaps because not many teams are willing to pay so much so early).

So startup founders are forced to choose whether to bite the bullet with AWS, or to move fast and pay a premium for tools like Firebase — only to have to rebuild from scratch later anyway. Isn’t that ridiculous?

# There must be a better way

At Digger we believe that having to figure security groups, access controls and networking before your app can see the light of day is plain wrong. It should be the other way around — go live first, fine tune later. We believe in meaningful defaults, not hard limitations. Complexity should be optional, but still accessible if needed.

We are building a radically better developer experience for your cloud infrastructure. Instead of locking you into an oversimplified model like Heroku we manage your AWS account and give your team self-service tools to launch new stacks, manage environments and deploy your apps and services. Engineers can move as fast as in the early days while also retaining full control over the internals with Terraform and AWS CLI.

If you ever worked for a big tech company you probably know what I’m talking about, because most of them have built in-house tooling on top of their AWS, Google or Azure setups. We certainly did at Palantir, Amazon and Fitbit — without it we would barely ship anything. We had to build it, otherwise developers and DevOps engineers would block each other all the time. But there is no reason why only big tech should benefit from this kind of next-generation tooling; so we left our jobs and started Digger.

One may ask, aren’t things like Elastic Beanstalk and Google App Engine solving this problem? Not really. On paper maybe, but in practice they are worse than Heroku — expensive, limiting and you have to rebuild from scratch eventually. Kubernetes, managed containers and Serverless are around for ages in every cloud provider, and yet if you want a modern future-proof stack the only way so far has been the hard way. Not anymore!

It is still early days for us. Our small team is based out of London, we are backed by EF. We strongly believe that the developer experience gap is the industry bottleneck today, and bridging it will have transformative impact.

If this excites you, do get in touch. We are looking for early adopters as well as to grow the team.

https://redd.it/k9f8hw
@r_devops

Circle CI caching

Anybody familiar with CircleCI and how I can cache my repository so I don't have to pull it everytime?

https://redd.it/k8spgt
@r_devops

[Newbie Alert] How and where to start with devops?

Hello people, I need your help how to learn devops skills.

I've been developer for 5 years and I see that my life would be easier if I can automate bunch of things and learn how to use tools like Ansible, Docker, Kubernetes, etc. So, I need your help and advise where to start.
I am indecisive should I start with linux administration or with docker or something else? I don't know what exact steps I need to take and I do not have the road map that I will follow in order to learn and become devops engineer.


Can you please help me with this problem?
thanks

https://redd.it/k9prbw
@r_devops

Advice: Overcoming AWS Cloudwatch Alarm limitations

tldr; Cloudwatch Alarms requiring individual instance alarm set up is a no-go and I'm trying to find a good way to get cross-region, large volume, metric based alerting going and so far the best solution I have is Grafana.

---

So I'm running into an issue with Cloudwatch Alarms where they really aren't quite matching up to what I would want out of an enterprise cloud alert system.

Essentially, the need is: Alarm on x (unknown) instances based on violations of Cloudwatch metrics.

Rather than be able to set a threshold and then alert for any instance that violates that threshold, it looks like instead you are required to set an individual alarm for each instance you wish to monitor. This is highly, highly ineffective for me as I am monitoring a large number of instances cross-region. I looked into "composite alarms" but from what I understand, composite alarms are really meant for a binary yes/no is something in violation, and you lose the detail of which instance is having problems, and what problems they are having.

Without ranting too much about pet architecture alarms in a cattle environment, I have been trying to find some alternatives and the best I've come up with so far is using Grafana to poll metrics from Cloudwatch and then set up the alerts and notifications from Grafana.

Is there potentially a better strategy I'm overlooking here? I'm honestly a bit baffled that this is the way that Cloudwatch Alarms work as well, so if I'm missing something here, please let me know.

Side note: Using terraform to standardize infra with alarms as part of the resource definition has been tabled- so that is unfortunately not an available option for me.

https://redd.it/k9qwiv
@r_devops

‏ Any good engineering learning app/YT channel to listen to while driving?


I'm driving around 35 minutes to and from work each day, is there any good app/youtube channel to listen to while driving to learn from?

Topics: Networks, Product management, System engineering, DEVOPS and so on..

https://redd.it/k9q7cj
@r_devops

What is your experience of working with ITIL practitioners?

Hey everyone.

I read a short article about why DevOps engineers should collaborate with ITIL practitioners ([here](https://exceed-team.com/tech/how-to-align-devops-and-itsm)) and I'd like to know your opinion about this approach because I considered that DevOps in closer to Agile methodology, and ITIL is quite old.

So what do you think? Maybe you can share your experience?

https://redd.it/k9qilr
@r_devops

Continuous Delivery in the Age of Microservices and COVID-19 - Interview with Tracy Miranda, CD Foundation, Exec Dir

In this discussion with Tracy Miranda, Executive Director of the Continuous Delivery Foundation, we talked about some of the pain points the organizations face when bolstering their CD practices and how the Foundation is helping to address them.

[Continuous Delivery in the Age of Microservices and COVID-19 - Interview with Tracy Miranda, CD Foundation, Exec Dir](https://www.linux.com/audience/devops/continuous-delivery-in-the-age-of-microservices-and-covid-19/)

https://redd.it/k9tstr
@r_devops

so many tools and frameworks in the DevOps Ecosystem does anybody build AddOn or Complementary tool to get additional income?

Hello
The DevOps Ecosystem is so big, looking for examples for paid plugins or Complementary tools
That Get to the creators some revenue.
I'm talking about indie developers, not big established companies
any examples?

https://redd.it/k9t2qt
@r_devops

Considering using AWS Copilot for production stack. Is it a good or bad idea?

AWS made [Copilot](https://aws.github.io/copilot-cli/) generally available recently and it looks so cool. Like, the simplicity of things like Firebase and Vercel but in AWS so future-proof and won't get unreasonably expensive.


Wondering what could be the reasons against using it for production? Has anyone used it already? Are there any significant tradeoffs or limitations?

https://redd.it/k9uk77
@r_devops

I am just not understanding Devops

I am a sys admin and always have been. I script tasks to make them easier, and I like to think my servers are cattle not pets. However, in my job we simply stand up servers and services. I provide AD, email, Sharepoint, various other services. But we don't develop any apps in house. I don't feel like I have an opportunity to do any of the "Dev" parts of Devops, and don't really understand how to begin. Even if I start to utilize my homelab, I'm not a programmer, so I don't really have anything to deploy. I don't have anything to use with CI/CD, or to place in a container. How does someone who is only a sysadmin, who only works as a sysadmin, even begin to practice Devops skills when it never comes up in their professional life?

https://redd.it/k9x7gv
@r_devops

Any automation before Pull Request?

Upon Pull Request...

Do you run automation and then notify the reviewer that it’s “all clear” to do the code review?

Or, do you just let the reviewer review based on the PR and then if it’s approved/merged, then the automation happens?

I guess I’m trying to figure out if anyone thinks it’s overkill or pointless to do any automation before the code review.

https://redd.it/k9xpcf
@r_devops

GCP Memorystore Redis queueing with GKE question

Hey, anyone using google managed redis instances in Memorystore and connecting from GKE?

I've got a built out webapp in gke already, and an existing redis instance in memorystore. Just curious on how to best connect then two to be a queue management system. Havent used redis enough so if someone could assist would be cool. Thanks in advance

https://redd.it/k9vlye
@r_devops

Helm Charts with Openshift

Hey guys !

I’m new to the Ops side of the DevOps thing. I am a developer and have joined a small company after leaving a big one. Because of my experience working in a DevOps environment, they want me to build them some things. I was able to build out couple jobs in Jenkins and was also able to containerize some apps.

But now, the ask is that I need to use helm charts with Openshift to deploy these docker containers and I can’t seem to find good tutorials for the both. I can find tutorials for Openshift and for Helm but not for both.

And that’s brings me here. I’d be grateful if someone can point me to a nice tutorial ? I have experience with Openshift web console and CLI from a developer stand point but no experience what so ever with Helm.

https://redd.it/k9xrmk
@r_devops

Gitlab job exiting error code 1

Hello everyone !

So ... I'm having a problem running a job inside gitlab. I'm using it as a docker container, which both he and the runner are running on a server with CentOS 8 as the main OS. Below I will put information about the job log and the .gitlab-ci that I am using to try to build the application.

[Job log](https://pastebin.com/HvTS1L1c)

[runner log](https://pastebin.com/ugxW4NvH)

[.gitlab-ci.yml](https://pastebin.com/GtbQ9EaX)

Docker version: 19.03.13, build 4484c46d9d

Gitlab version: 13.4.1 (c90be62bdef)

Runner version: 13.4.1

Thanks in advance for who can help me with this issue.

https://redd.it/k9qop2
@r_devops

if you're a web developer please don't ignore this question.



hey in this year I have learned (MERN) stack and joined at least 5 internships in 4 months, in all of them for the first week I worked very hard and try to do all the task's before deadline's but in the second week I get very slow I feel like my mind is not working and I miss every deadline and end up disappointing my senior's I feel like I am dragging the team back and eventually I quite every one of my internship I have had, I never got fired I just quit, is it performance anxiety or something else I don't know but I know that I like codding but when someone gives me the task or when I see that the deadline is too narrow for just the simple task I panicked and I feel I am not able to do that task can you tell your experience is that ever happened to you.

I could have asked this question on Quora but I thought this is the best place to get this particular question answer.

Thank

https://redd.it/k9omlf
@r_devops

Pluralsight or acloudguru

Which one is preferred for devops related trainings. Please suggest if there are any other alternatives. My organization will be paying for it. They are suggesting pluralsight, but I heard that many courses of pluralsight are outdated.

https://redd.it/kabtmg
@r_devops

AWS Compliance scan with cloudquery

I've added compliance queries for AWS - [https://github.com/cloudquery/cloudquery#aws-compliance-pack](https://github.com/cloudquery/cloudquery#aws-compliance-pack) . I'd love to get feedback

https://redd.it/kaec29
@r_devops

What's the best setup for an on-prem server running a handful of containers?

I am moving some of our internal tools out of the cloud and into a rack in my office. Today I am setting up a new server to run a handful of webapps, most likely in containers.

The plan is to setup oauth2-proxy and then a handful of low-traffic, internal use only, mostly open source apps.

I would like to be able to update the apps independently, and to be able to tell at a glance what all is running and if they are good. We are a small team and these tools are nice to have a more than mission critical. So I don't need a bunch of residency as long as I can recreate the setup relatively easily.

These containers will be the only thing running on this server. The database is on another server and will be accessed via an isolated network (public Nic for proxy, internal Nic db/admin).

I don't have much experience with kubernetes, and that seems to be overkill. My current thinking is that docker-compose could be a good fit, especially since I could ensure the proxy and apps are configured in the same place.

But I don't know what I don't know.

Is there a distro that really excels at something like this? Or maybe some kind of admin panel (cockpit?) that makes managing a single container server a breeze?

https://redd.it/kaeosn
@r_devops

Noob Question - Can(or should) I Orchestrate Containers with Different Configurations?

So I've started a business, and we need a way to quickly deploy very similar applications. In essence they're simple web applications that hook into some niche data, and can quickly provide websites. Each container will contain one instance of the application(of course more can be added depending on the client's needs), but each instance of the application is tailored a bit to the customer. The number of containers will get pretty high so I feel that container orchestration is definitely necessary to easily keep the application up to date. In my mind, Kubernetes will work just fine in this situation I'm just not too sure.

https://redd.it/kahnt8
@r_devops