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
Kubernetes: tracing requests with AWS X-Ray, and Grafana data source is a step-by-step guide on how to setup tracing in your EKS cluster using AWS X-Ray by Arseniy Zinchenko - a member of the Ukrainian DevOps community.

Also, make sure to subscribe to his Substack! He posts new things quite often and I have no idea where does he find time and willpower to do so 😅

#aws #kubernetes #observability
👍101
​​I got a bit distracted in the recent days, so I make posts with delays.

Today we have a Donations Monday with a twist.

We are raising funds for two foundations at the same time:

- For NayTak for camouflage nets.
- For UA Responders an IVL and a defibrillator for medics from Kraken.

The twist is that you can win a remnant of an S-300 rocket (on the picture).

Every donation for >50 UAH is a chance to win!

You can donate on:

- a Monobank jar: https://send.monobank.ua/jar/5SizeGGzBM
- top up the card directly: 5375 4112 1191 0851

Please, add your contact details if you don’t use MonoBank for donations, so they know how to find you in case you win.

P.S. Tomorrow I will send a new newsletter issue, that I should’ve sent yesterday.

#Donations #Ukraine
7👍3🔥1💩1
I know that many folks have a mix of corporate and private repositories on their laptops. By private, I mean their small projects, dotfiles repo, forks of public repositories, etc.

Here's an interesting guide on one of the ways of how to keep separate users (email, signing key, name) for different repositories.

This would also work if you have repositories backed by different provides. For example, if you need to push to both GitHub and GitLab and use different SSH keys for that.

Here's a condensed version of this article on StackOverflow.

#git
👍9
​​On behalf of the Architecture Stage organizational committee I want to invite you to the DOU Day Conference!

It’ll take place offline in Kyiv on the 18th of May.

The thing is that if you buy a ticket now, you will get -50% off for the second one. So, a great opportunity to grab some tickets together with your teammates.

#event
👍2🤔1
The core idea of this article is pretty simple: you need to protect your Terraform states. I don't think this is a debatable topic, and anyone has a different opinion on this matter.

To quote the article itself:

 an attacker can modify the Terraform state file it’s game over and bad times ahead.


However, in the very end, this article provides some suggestions that I never saw implemented IRL:

- Store the state lock in a separately permissioned location
- Use a read-only role for terraform plan executions

#terraform
👍4
​​For todays Donations Monday I want to share with you a fundraiser for 101 Starlink terminals by Dzyga Paw foundation:

https://dzygaspaw.com/starlinks-101

#Donations #Ukraine
👍7💩1
Some time ago, I predicted that there is going to be more Kubernetes distributions. Then it didn’t happen, so I thought I was wrong.

Yet, now Canonical has introduced their Kubernetes distribution.

It’s based on the upstream Kubernetes 1.30, has some built-in add-ons, and yes, you can install it with snap.

#kubernetes
🔥9👍1😢1🫡1
A neat comparison between Argo and Flux.

I like the fact that the article mostly focuses on UX and use cases. Yet, I would disagree with this statement here:
 would deploy one Argo CD per tenant, where each tenant is an independent developer team with their applications, but it can work with multiple clusters, for example, dev/stage/prod, etc.

I mean sure. If you have enough resources, go for it! However, I witnessed how getting down from “Argo per team” to “just one Argo” reduced resource consumption in a cluster by 90%.

#cicd #gitops
👍8👎3🔥2😱2
People often say that Observability is a Data problem. Although, it sounds correct intuitively, I cannot say that I fully understood how Data engineering approaches could be applied to the Observability systems.

This article about Wide Events clarified things for me a bit. Indeed, if any event that happened in the system is just an object with some value and useful metadata, things like metrics, logs, and traces become less relevant - it's all events now!

Apparently, this is how Observability is done in Meta, according to the author, and apparently people in Meta like it. I never worked for Meta, I don't know what they really use there and if it's better than the tools available to us mere mortals.

However, this is an interesting concept, and it would be wonderful to see similar projects that are not internal to the Big Tech companies.

#observability
👎41👍1
Kondense is a Kubernetes tool that allows you resize contianers in a pod based on the memory pressure.

It’s installed as a sidecar and uses real-time memory pressure to determine the optimal memory for each containers in a pod.

You can read the justification behind this tool in this Reddit post

#kubernetes
🔥3👍1
RedHat reported a 10/10 vulnerability in the xz compression library.

The vulnerability provides remote backdoor access and present in xz 5.6.0 and 5.6.1.

There’s also an interesting discussion of this vulnerability on HackerNews:
 annoying - the apparent author of the backdoor was in communication with me over several weeks trying to get xz 5.6.x added to Fedora 40 & 41 because of its "great new features".

#security
😱16👍2