Roles in DevOps team?

What are the required roles in a DevOps team? In your company, are these roles done by unique people, or do you assign individuals multiple roles?

Also, is there a minimum amount of qualified people to have for a DevOps team? I'm thinking 2 qualified as a basic minimum to have a backup.

https://redd.it/s9bwta
@r_devops

Devops for batch scripts

Hi,
Our applications are running using batch scripts on control-m. Different batch scripts either call .net exe, ssis package or db proc, sql script or batch commands.
We want to implement devops for same.
My questions are
1. Should the batch as well as exe dotnet code shoild be together and ms build creates the exe?
2. We have 100+ control m jobs which calls batch files. Do we need to create 100 jobs in jenkins to create deploent bits and then 100 CD jobs for deployment.

Any link to understand how can implement devops for batch scripgs is appreciated

Thanks

https://redd.it/s99w62
@r_devops

How do I interview a DevOps engineer?

When hiring software engineers, I typically have them demonstrate something in a language they'd be using on a team. How do I effectively get a DevOps engineer to demonstrate their skills in a 1 hour interview? I just don't know what is common out there. System design? Hands on docker / kubectl? What have you found effective?

https://redd.it/s9eh2i
@r_devops

A Primer on the History and Evolution of Incident Management to Today

we thought itd be interesting to write about the history of incident management how concepts like ICS and NIMS were born

https://redd.it/s9fgro
@r_devops

Change control management for psuedo devops / cloud infra team

TLDR:

Looking for help from the community on how our team can navigate the change control process while still being agile and not being blocked. Our team manages the aws cloud infra of the non engineering team (all of IT related infra) and we also handle automation via python lambda functions / building apps in ecs/fargate. We don’t manage pipeline of engineering, but we do manage our own pipelines for our code base.

This compounds when we get past the theoretical and into reality: sometimes when working on a task, as a developer/infra engineer you realize that there are other dependencies that need to be configured/built in order to complete your task. In our current situation, that would require yet another change control request that may take days to get approved. This just simply isn’t scalable

Our goal is to maintain a way to manage production without having to wait up to a week to actually execute our code/terraform changes. How are you all handling the rift between productivity/speed against rigid change control processes?

Details:

So, our company has exploded in growth and there has been a new directive coming down from upper management to make no changes to production without a change control request that is approved by four different teams. Including break/fix. There is a change control meeting once a week where teams have the ability to get approval for production changes, and we also have emergency change controls that we can submit but only for emergency break/fix.

This new directive is extremely broad; the problem is that basically everything is considered production. This means that we effectively have to wait days or sometimes a full week to get work done (something as simple as changing a comment in our code). If we need to troubleshoot something in real time, we have to get a CR for that. IMO this isn’t scalable and is an indication of a collision between traditional IT and modern day cloud management.

Unfortunately we don’t have true dev/prod environment segregation, and the ones that we do have don’t have real parity with production.

How is everyone here handling this situation? Any advice on how I can convince management to allow us to use our professional discretion?

The only real solution I can think of is to leverage emergency Change control requests on a daily basis, however this will undoubtedly be shot down by management.

https://redd.it/s9h50i
@r_devops

How often do you all meet with your managers to discuss job performance? What topics do you typically go over during these meetings?

At my last role I met with my manager a grand total of one time 5 months in and now realize this likely isn't the norm, so I'm curious how often you all meet with your manager to discuss job performance, and what topics you typically cover during the meeting? I realize this isn't directly about devops itself but I do work in a devops context so that's why I'm posting here.

https://redd.it/s9ilod
@r_devops

New Relic's Helm chart repository appears offline

Just a heads up in case you're seeing similar woes -- my CD pipelines are trying to install from https://helm-charts.newrelic.com, which is currently 404.

Here's hoping New Relic gets their stuff fixed!

https://redd.it/s9hsrb
@r_devops

Production level AKS cluster using Terraform

Hello people,

I've been exploring Terraform script to create an AKS Cluster with multiple subnets, Ingress and Egress rules.
And other extra features as what will be required for production usage. And won't seem to find anything great.

Can anyone please help on this.

Thanks.

https://redd.it/s9htbu
@r_devops

Jenkins vs Circle CI vs Atlassian Bamboo vs Others

Hey everyone, thanks for taking some time to read this, as this might be a bit of a long one.

So currently we have one older instance of Jenkins we have running, and myself coming in and introducing a number of DevOps concepts and having some say with web, we got another instance of Jenkins running to start off "clean". I had experience (all be it several years ago) with jenkins, and I was comfortable working with it. Getting use to the new declarative pipelines took a bit, as much needed to be done with script blocks, conditionals and so on.

Well I've been reading up about how Jenkins feels antiquated, and recently had an experience of a few plugin upgrades (much of it the core aspects of Jenkins, with a few additional plugins (like notification to bitbucket for example), where it just totally broke pipelines. Talking with a few others this seems to be a regular occurrence.

For examples in our repos that are git based, they often use sub-modules, well found out during the checkSCM this doesnt pull in submodules, so I wrote a script to ensure that they are initialized during pulls and we check out the right branch in those submodules. Well that totally broke randomly after doing updates, where script blocks no longer worked as expected. It took me about a day and a half to fix things, and even then they are still not quiet working.

So this is all boiling down into....should we keep using Jenkins when we need to be able to reliably update things, and know in doing updates our scripts dont break while giving us the functionality we need?

I've been looking into CircleCI and Bamboo because I know both of those can be deployed locally, which is very important for us. But I was wondering what are others experiences with these, or would you recommend something different? Like things you like or dont like about it, quirks that are just aggravating?

Some things we are looking for (which should be standard):

\- Smooth integration with git
\- Logging and reporting of errors
\- Custom pipelines that allow scripting
\- Some ease of integrating systems together
\- Being able to use 3rd party programs (ansible, terraform, etc)

Would appreciate any opinions or share of your own experiences.

https://redd.it/s9mh05
@r_devops

HIRING Senior DevOps Engineer REMOTE

I’m \#hiring. Know anyone who might be interested?


We are experiencing rapid growth and looking to bolster our talented team by adding skilled engineers.


Apply now and start a journey to your next professional experience.


Apply here: https://www.linkedin.com/jobs/view/2888940357

https://redd.it/s9nx6d
@r_devops

How should I deploy an app that is dependent on third party services like Meilisearch?

Hello, my first post in this subreddit.

I'm more of a developer but this time I need to know the most efficient way to deploy a Laravel app that has docker-compose inside with 2 services defined: meilisearch and redis. If it was only for Laravel app itself, then it would be straightforward: one web service. I'm using Digital Ocean for deployment. More specifically App Platform feature because it is a PaaS. I don't have to focus on server infrastructure so much. You can compare this feature to Heroku.

However, a service like Meilisearch has it's own url and port in development environment like https://localhost:7700. So I guess besides one web service for Laravel app itself, I need to create another one for Meilisearch and connect the app to this? That would be two times more expensive. Is there a better way to handle these things without being more expensive? is it better to use Droplets instead? Then it's more towards IaaS I suppose? Or what kind of deployment is considered as 'industry standard' or in your opinion 'most efficient' when you use docker-compose with third party services that connect with your app in development environment?

https://redd.it/s9ooi8
@r_devops

Production level AKS cluster using Terraform

Hello people,

I've been exploring Terraform script to create an AKS Cluster with multiple subnets, Ingress and Egress rules.
And other extra features as what will be required for production usage. And won't seem to find anything great.

Can anyone please help on this.

Thanks.

https://redd.it/s9htbu
@r_devops

Question for you smart people about Jenkins and Blue Ocean.

I want my users to only use blue ocean because that’s what I’ve been told to do. That said, I want to remove the top right link that puts the gui back to classic view. Is that possible? Anyone ever do that? Thank you in advance.

https://redd.it/s9rs0u
@r_devops

Will I fall behind in my career if a company does not use Kubernetes?

If I join a big data company that primarily uses serverless and various AWS managed services for data engineering like Redshift, Elastic MapReduce, Lambda, etc, will I be less marketable than an engineer who has Kubernetes and container experience?

https://redd.it/s9o199
@r_devops

How do I interview as a DevOps candidate?

I feel pretty confident about my ability to present myself as an employable candidate, but I would ask, what are common questions you have for companies when you interview to size them up?

* Tooling?
* Compensation?
* On call expectations?
* Work/life balance?
* WFH availability?
* Scope of responsibility?
* Competing companies?


What did I miss?

https://redd.it/s9toik
@r_devops

Interviewed at potential dream job

Seriously, wowzers. Job has it all, Culture, pay, time off, retirement... you name it!


They had 280 applicants and the first interview was a culture fit interview and I nailed it.
They said that it could take weeks to finish the process, since they had a large volume of applicants.
I also have a great interview on Monday that they've already explained I will probably get.



Should I wait if offered? how long should I wait? Both jobs seem great, but the one I mentioned above is clearly much better.


I'm nervous.

https://redd.it/s9iw0z
@r_devops

Appreciate some feedback

Would appreciate some feedback on the article I have written folks. Thank you for spending time.

​

https://www.linkedin.com/posts/idushyantmishra\_devtools-gtm-buildfromindia-activity-6890270010779021312-q5yA

https://redd.it/s9fyjj
@r_devops

Secure AND unsupervised Developer Machine

Hi there,

How can we have a unsupervised/ locked macbook for the developers so they can install what they feel like using to be as happy and productive as they want and AT THE SAME TIME make the network, Dev, Stg, and PRD environments as secure as possible?

I work for a financial institution and I am friendly challenging the way they limit what can be done and installed into the laptops. So I am looking for a better way to suggest them.

Any help would be appreciated.

Thanks

https://redd.it/s98v6t
@r_devops

Centralized interface for managing supervisor processes

Hi, i currently manage several bodies that run services through supervisor. To restart processes through supervisorctl I do it through ssh within the instance.

I’m looking for a centralized interface to manage all supervisory processes in different instances.

I’ve tried tools like CeSi but I get a lot of Python errors and so I see the interface is no longer supported.


Any suggestions? Regards

https://redd.it/s99sxc
@r_devops

Guided DevOps meditation

Hope this provides some much needed stress relief

https://youtu.be/epcbx5HkCbM

https://redd.it/s9drtn
@r_devops

Free CMDB tools integrated with Jira

Hello everyone,
I'm looking for free and open source CMDB solutions that we can integrate with jira ?
Insight-asset management, is an atlassian CMDB solution, integrated with jira but just on premium and enterprise version.
If someone has an idea about those free solutions and how to integrate them with jira on both ways, please write me here.
Thank you.

https://redd.it/s8ubiq
@r_devops