CatOps – Telegram
CatOps
5.09K subscribers
94 photos
5 videos
19 files
2.57K links
DevOps and other issues by Yurii Rochniak (@grem1in) - SRE @ Preply && Maksym Vlasov (@MaxymVlasov) - Engineer @ Star. Opinions on our own.

We do not post ads including event announcements. Please, do not bother us with such requests!
Download Telegram
Forwarded from Sysadmin Tools 🇺🇦
Обновите свой git, пжалста

https://www.openwall.com/lists/oss-security/2021/03/09/3

#git #exploit
Linux Foundation CKA cource:
Due to the difficulty in pronouncing the [Kubernetes] name, many will use a nickname, K8s, as Kubernetes has eight letters between K and S. The nickname is pronounced like Kate's.
How you usually call K8s? (Multiple answers)
Anonymous Poll
7%
KubEnetes
50%
KubernEtes
7%
Kate's
7%
Key Eight Es
52%
Kuber
6%
That orchestration stuff, you know
14%
Oh no, again...
5%
Other
Some good practices on how to tag your AWS infrastructure for better cost management and compliance audits.

As someone, who works with AWS on the daily basis, I can say that this article is pretty useful. We are trying to create a similar account structure / tagging in the company I work for as well.

This article sums up ideas regarding proper resources tagging and also AWS account layout, which can help you to organize and track your assets better.

It recommends some third-party tools for cost audits as well.

#aws
An article on how to scale Celery workers in Kubernetes based on RabbitMQ queue depth.

Basically, this is an example of how to use custom metrics for scaling. They were using KEDA to collect those metrics. So, it may also be interesting to those who want to know more about KEDA itself.

#kubernetes
"Quick read" by John Arundel called Writing slower Go programs. Actually, this can relate not only to Go.

The idea is that you'd better optimize your code for readability, brevity, and simplicity to understand; rather than speed. Unless, you're working in an industry that requires blazing fast response times or works under shortage of resources. Such as game development, IoT, network equipment and so on. However, if you do work in such an industry, you likely know, what's you're doing.

#go #programming
At some point senior engineers may ask themselves: where do grow next? Is being a manager an only way to go?

That's not quite true. StaffEng has some guides as well as personal stories of people who grew beyond a Senior noscript, while still staying on the tech track.

#culture
Some best practices for GKE networking by, well, Google.

As was said in one book:unless you've actually done the work, you're in no position to encode it as a best practice.

#gcp #gke #kubernetes #networking
A new article by Julia Evans Get better at programming by learning how things work.

It’s impossible for a single person to know everything. Especially, in such complicated systems as we have in modern IT. And it’s fine, you can still do a good job.

However, you can become even better engineer by getting down to the nature of things. Especially, when you have a tricky bug to chase.

Here are some tips from Julia on how to get there:
- Just learning a few facts can help a lot
- Connect new facts to information you already know
- Ask yes/no questions
- Googling is a skill

#culture
This is not really a DevOps-ish article. However, salary topic is something that ususally interesting to everyone.

Here is an outcome of a personal research regarding salaries in the Netherlands and Europe. Personal in this case means that this research was primary driven by a private person.

I'm not familiar with dutch market personally, but overall the ideas described in this blog post are relevant to Germany as well.

Of course, things are quite different for non-EU/non-US countries. Especially, Post-soviet countries. However, it's a huge benefit that as an IT specialist you can compete not only on the local market, but also globally.
<for our russian-speaking subscribers>

Many thanks to “Потестим в проде” podcast for having me as a guest!

If you want to listen to me speaking about the common problems in devops-ish terminology, differences between IT market in Kyiv and Berlin, and if you need Kubernetes to run your personal blog, you can find the recordings on:

- YouTube
- Apple Podcasts
- Google Podcasts

Podcast is in Russian.

#slides #podcast
Thinking of starting a new open source project? Or your company wants to open some internal work?

At some point you'll need to start thinking about the licenses. Taking care of it earlier helps you to avoid problems like the recent issue of Rails and mimemagic

However, you're an engineer and not a lawyer, right? Here are just a couple of resources, which can help you navigate through the differences of the Open Source licenses.

- Choose an open source license - very brief and clear guide for beginners
- TLDRLegal - another quick start guide, but contains more types of licenses

#legal #foss
​​strace is a very old yet popular tool for Linux troubleshooting. Moreover, on many Ops interviews an interviewer is only satisfied when a candidate mentiones strace during the troubleshooting questions section. And of course, there are more bonus points for mentioning that strace shows system calls of a given process(es) and their arguments and return values.

Although, what are we trying to solve with it? And when to use it?

Julia Evans collected 9 common categories of problems which people use to solve with strace:

- where’s the config file?
- what other files does this program depend on?
- why is this program hanging?
- is this program stuck?
- why is this program slow?
- hidden permissions errors
- what command line arguments are being used?
- why is this network connection failing?
- why does this program succeed when run one way and fail when run in another way?

P.S. She also has a free webzine on how strace works!
A little story about how to drill down your AWS Costs using AWS Cost Explorer and some ad-hoc automation.

Might be interesting to those, who just has just started working with AWS Cost center or never bothered about $$$ before

#aws
We don't post any advertisements usually. However, I have my personal interest in this one.

As some of you may know, I'm part of the DevOps Days Kyiv conference committee. And we are going to have our next event very soon! It's going to be an online event, so you can join from any place and also it's totally free!

What: DevOps Days Kyiv 2021

Where: Online

When: 20-22 of April 2021

About: Top speakers from around the globe! You can check the whole list of upcoming talks on https://devopsdays.com.ua/

Also, we are going to have a fireside chat with one of the first Kubernetes contributors - Joe Beda! If you want to ask him anything, you can submit your question via this form: https://forms.gle/BghSK7Scobra6FDc8

Price: Free!

See you online at DevOps Days Kyiv 2021!
#event
AWS serverless patterns by Serverlessland.

Basically, this is a collection of common architecture patterns i.e. API Gateway to Lambda, Lambda and DynamoDB, API Gateway to Step Functions, etc. However, these examples also contain AWS SAM or AWS CDK templates, which you can use.

You can submit your own template as well. But no Terraform, sorry.

#aws #serverless
I missed this somehow, but now Kafka can work without Zookeeper! And I think these are great news!

New quorum controller, which is embeded in Kafka itself leverages Raft protocol under the hood. It will be added into the upcoming 2.8 release. However, it's already available in a preview mode.

Confluent managed to shave off some time for controlled shutdowns and the recovery time for uncontrolled shutdowns as well.

#kafka
A bit of original content for you today.

After I heard the news that Apache Mesos is about to be moved to attic I decided to make this short write up on why do I think that Kubernetes nowadays is not quite a competitor for other existing orchestrators right now. Well, it is, but there is room for those as well.

P.S. I also encourage you to read benjamin_mahler's comment to the news about Mesos. It provides some interesting historical perspective.

#original #kubernetes
Terraform 0.15 is here. Via the link is the upgrade guide from the version 0.14

Hopefully, no big code chages are required. Moreover, I already got a PR for 0.15 for my public module.

Also, Mitchell Hashimoto has mentioned in his Twitter that this version is a pre-release for Terraform 1.0. So, would the next on be 1.0?

#terraform #hashicorp