The Best DevOps Tools, Platforms, And Services In 2023?

[@DevOpsToolkit]

Fabio was very kind to create a wrap-up and personal observations from this video. Please go and check it out at github.com/FabsMuller/Top_DevOps_Tools_2023.

[@antoniogomeznavarro8095]

I just discovered you and I think you are the best DevOps/Platform Engineering channel in youtube, you’ll eventually get the recognition you deserve. Keep it up!

[@JOJO-oe7og]

Woow best DevOps channel yet! Bravoo

[@yotu9670]

Really love your videos. Straight on the point. Without adding any weird stuff. TAHNKS A LOT ❤

[@todddensmore4320]

Thanks!

[@DevOpsToolkit]

Thanks a ton

[@Jarek.]

These were best 30 minutes I invested in 2023 ❤ Many thanks for all these hints. My personal winner in Environments category is Kind (Kubernetes in Docker), basically - run a k8s inside docker. Awesome 😊

[@dyegodinizsouza]

Hi, devops here from Brazil. Thank you a lot for this video I really had the opportunity to learn more other tools that I'm not using yet.

[@CivoCloud]

Thank you so much for the first spot in the Cloud Computing category Viktor, that means a lot to us!

[@christi4nbellet]

You deserve it :)

[@swissrock1492]

I was looking forward to your 2023 lineup. Victor, you’re channel is the GOT when it comes to DevOps!

[@DavidBerglund]

Scaleway is the only viable/legal option (from this selection) to orgs within the EU that can't rely on cloud companies outside the EU (both referring to the location of the datacenters and company ownership). So Scaleway's so called "handicap" is exactly what makes it a winner for many EU orgs. I still think this is a great video btw! I would like to see more content (here or elsewhere) that is focused on the needs of orgs that can't rely on some managed cloud services in for compliance/security reasons. Self hosted solutions takes a lot of resources that most orgs don't have while still needing to adapt DevOps practices (which IMO demands the adoption of some cloud native practices).

[@mdaverde]

Fun video to watch. So many ideas in the Kubernetes space and it only keeps increasing!

[@ama054000]

Thanks for yet another great video Viktor! I came across a couple of tools over the past year or so while researching "KRM functions", specifically for Kustomize's implementation. I love Kustomize, KRM functions piqued my interest, and this led me down a rabbit hole. 1) I first stumbled acrosss something called "kpt". I would love to see a future video investigating kpt, which to me seems to be a potential replacement or competitor to both Kustomize and Helm. So perhaps by 2024 it could make the "App Management - Manifests" category mentioned in this video. 2) Google (?) also seemed to silently release their own CD application, "config-sync", which would likely attempt to compete with ArgoCD/flux. Would be nice to see an investigation into this.

[@DevOpsToolkit]

Kpt Is potentially awesome but still green. It might become a force to reckon. kzbin.info/www/bejne/jZLZmpqvppyBmrM

[@ama054000]

@@DevOpsToolkit 😅 I am embarassed -- I was "sure" that you had not made a video on this!

[@IvanRizzante]

Thanks for another great video! I would really love having videos on Loki, Tempo and Victoria Metrics!

[@DevOpsToolkit]

Adding those to my TODO list...

[@IvanRizzante]

@@DevOpsToolkit thank you! I just wanted to add that Victoria metrics didn't support Prometheus exemplars at least at the time when I last checked it. I think exemplars are very useful for investigating anomalies and it's a pity not to have them in their software

[@DevOpsToolkit]

@IvanRizzante that true. I missed that one.

[@camaycama7479]

I would be fantastic if you could do a video with a live use case of using Harvester along with Rancher. Could be cool to see the Harvester VM new paradigm in action. Thank you!

[@DevOpsToolkit]

Adding it to my to-do list...

[@ioannisgko]

Great list of tools, I would only add that the Jenkins helm chart is excellent, it includes the Kubernetes plugin for Jenkins, we can define specific plugins to be installed at startup (see JCasC) and do a lot more. I really like that the Kubernetes plugin allows us to define container images for each stage, and the agents are pods that get created and destroyed automatically. So, there is no need to configure or manage Jenkins agents, it's all done automagically 🙂

[@christopherrichard6567]

Had an actual LOL when you described AWS/GCP/Azure as 'doping'

[@dtanuj11]

I think you could have another category in Networking and Security as well. Since eBPF, WASM, servicemeshes are also what people are looking at and their are many to compare as well. As well as multiple vendors. What do you think?

[@DevOpsToolkit]

This will be the year that will mark the death of sidecars. I am preparing a lot of content on those subjects. It's coming soon.

[@gorgegorgara2186]

Hey Victor, how about JenkisX, it's a CloudNative CI/CD tool oriented surround ChatOps (commenting on pull request to trigger stuff). Underneath it uses Tekton for job execution and Lighthouse which is basically a fork of Prow (the CI/CD tool used for Kubernetes itself). I would love a video about it and how it compares to other solutions

[@DevOpsToolkit]

The last time I checked jx, it was abandoned. All the maintainers went to other projects and there was no activity. That was a while ago. I'm not sure what it's status is today. Did it get back to active development?

[@CrashTheGooner]

Ty Viktor I watched this lineup last year as well! Man u insipire me bro!

[@DevOpsToolkit]

Glad to hear it!

[@xyjlugah9852]

you missed Concourse in the CI/CD tools section. it's pretty popular too.

[@DevOpsToolkit]

I have not used it yet. Adding it to my todo list...

[@memo.i.garcia]

Man I love you, thanks for sharing!

[@mozhago8280]

Hi Viktor, Great Content as usual, I think you forgot the most important thing "Software supply chain"!

[@DevOpsToolkit]

Next year...

[@90DaysOfDevOps]

Great video and even for someone in the space and constantly learning your videos are fire! I now have a plan to learn using your content some of these areas. Would love to see more data focused apps and databases appearing next year.

[@DevOpsToolkit]

How about "everything you need for your app all the way until it is running in production (including DBs)"? Stay tuned if that sounds good since we're launching a new series at the beginning of February. I cannot say more at this time. Give me a few weeks to go public with the idea.

[@90DaysOfDevOps]

@@DevOpsToolkit like it. Day 2 ops for your apps.

[@FluidITGuy]

I'm looking for a head to head on a full platform security solution. Wiz, orca, sysdig, crowdstrike falcon and Aqua. I don't know whom to pick for a demo. I want one tool, not many for CNAPP

[@DevOpsToolkit]

I'll do my best to create a video about a security choices.

[@salborough2]

Awesome video - thanks so much Victor :)

[@valour.se47]

Cdk8s and all the finalists are awesome. Even though at my work place I am stuck with helm charts hell

[@markbaumgarten]

Great stuff! - Would love to see a similar video given the scenario of being on prem and using bare metal. I have my own ideas on potential winners but would love to see your best "olympic candidates"

[@DevOpsToolkit]

Adding it to my TODO list... Until then, most of the story is the same. The major difference is with the tools you use to setup and manage a Kubernetes cluster itself. What's happening inside it should be mostly the same. Now, if the situation is also air-gapped, than the story changes significantly since you would not be able to use any of the SaaS solutions.

[@markbaumgarten]

@@DevOpsToolkit Yes - I meant to include the air-gap challenge, but forgot. Sidero metal / Talos is my current #1 choice. Then bootstrap all required applications (ingress, rook-ceph, argo + custom apps) in the online cluster using a series of pull through image registries. This way all image layers are serialized to disk(and easy to copy offline). Then manage cluster apps using argocd+git repo from here. I noticed that in a video you created two years ago (automate everything using argo) that there is a bare minimum of manual setting up involved before argo can do its thing. Im wondering what the best way of getting that initial cluster up and running might be, given that today it has been 2 full years....

[@DevOpsToolkit]

@@markbaumgarten In airgapped, a minimum setup, besides Kubernetes (including networking and storage), would be Argo CD, an internal registry (e.g., Harbor), and a push of all the images and charts you need to that registry. From there on, you should be able to setup everything you need by pushing manifests to Git.

[@markbaumgarten]

@@DevOpsToolkit Thanks. I did hear about Harbor, but did not give it a go yet. On the Harbor docs it says: "Harbor only supports proxy caching for Docker Hub and Harbor registries." When deploying ...say... rook-ceph + ingress-nginx + argoXXX + etc, a whole myriad of registries are being used. Also having to tag + push all used images in my own testing/dev environment is cumbersome and means that I have to modify all of my manifests when deploying the same stuff offline....so I don´t really want to go down this path. Instead I deal with this by mirroring the image layers from each of my individual online registries and manipulating http + DNS on my offline cluster(s). Serving one pull through cache per registry I need online while serving the mirrored data as a normal registry offline.... Using this strategy I can deploy anything online and expect everything to appear/work in exactly the same manner - but air-gapped and not having to fiddle with manifests. All I have to copy is manifests and registry image layers. I am still on the lookout for a simpler solution, like Harbor, but one that allows any registry one would want to use. Still haven´t found it though.

[@JesusAlbertoSuarezMendez]

I like these tools. It would be fine to have The best security DevOps tools, with disciplines like secret protection, security scanning, access management, etc. Thanks

[@puyansude]

Great work, Thank You!

[@fabianoslack4269]

21:37 why gitlab is not in these cases?

[@DevOpsToolkit]

Partly because i don't use GitLab for my repos so I don't use GitLab CI either. I did not want to include tools that i haven't been using lately myself. There's a ton of other tools that could have been included and i think i said somewhere around the beginning of the video that i choose to restrict the scope to "i used those not long ago".

[@fabianoslack4269]

@@DevOpsToolkit thanks so much!!!

[@nnamdizobeashia7320]

Where can i learn devops and most tools used here, again thank you DevOps Toolkit

[@andrewnhien9714]

What do you think about software supply chain security? Please do the video about it!!! Many thanks!

[@DevOpsToolkit]

I have one on that subject in progress :)

[@chrisre2751]

Thank you. Great video

[@Subbeh2]

Amazing content! Especially for someone new to DevOps. That being said, what about tools like Ansible and Puppet? Are they not used much within DevOps?

[@DevOpsToolkit]

Both Ansible and Puppet are used heavily but their adoption is not increasing. They're in a similar situation like many other older tools that were dominant in the past and are now replaced by better tools, yet, it's too expensive to replace them so they're here to stay for a while longer. Heck, mainframe is still widely used. That being said, Ansible is still one of the best if not the best tool to manage on-prem infrastructure and applications not based on containers, as long as those are managed in mutable way. For immutable management of resources, Terraform took over from Ansible and was, later on, reinforced by Pulumi. Now something simiular is happening with Crossplane taking over Terraform and Pulumi dominance. As a side note, when Chef and Puppet emerged, many said that those will not replace CFEngine, yet, to an extent, they did just as Ansible took over the crown from Chef and Puppet. We advance technology by improving it but, over time, we often realize that improvements cannot get us (without too much work) where we want to be and, as a result, come up with new ones. The same cycle is happening with every tech.

[@Subbeh2]

@@DevOpsToolkit thanks so much for elaborating. Makes sense 👍🏻

[@rampanwar1316]

Great video. I have a question can we build docker image using Argo workflow ?

[@ioannisgko]

You can use Kaniko to build a container image within Argo Workflows

[@DevOpsToolkit]

Yes you can! Argok Workflows is a pipeline type of a tool that allows you to execute any set of commands. Now, if you need a container image built with Docker, you would need to connect your pipeline with Docker running somewhere (some server). If you do not have to use Docker for that (as you shouldn't), you can build container images with many other tools. You can, for example, use Kaniko. The important part to understand is that a container image is the same no matter which tool you used to build it.

[@rampanwar1316]

Thanks ❤

[@mbesida]

What about Datadog for dashboards?

[@DevOpsToolkit]

Typically, we use dd for everything related to observability not only dashboards. Dd is a great service.

[@Matej-f9m]

what about kpt and Backstage as application management tools?

[@DevOpsToolkit]

Both are great. It's just that I could not cover all the categories so I focused on those i was most involved with.

[@fpvclub7256]

Great video and list. I agree with most of your selections - I prefer Devspace and Helm .. but otherwise you nailed it :D

[@sergeygr]

Great video! Thank you! What do you think about Sentry?

[@DevOpsToolkit]

Unfortunately, I haven't used it yet. Adding it to my TODO list...

[@rjarora]

Shocked to see that CircleCI didn't make the finals!

[@msm1723]

@DevOpsToolkit Good day! First of all thank you for videos! Great job :) I am a QA and want to increase my understanding of devops stuff. I just finished couple courses on Coursera. However I am having really hard times sorting all those parts. As I understand this video is itself a good enumeration of all aspects that adds up to healthy devops environment. However I wish you make a video explaining each of this categories in simple words, maybe even draw a schema of relations between those parts. Or maybe you could advise some source of this secret knowledge? :) Anyway thanks again! Will try to go through all this categories, understanding what is going on inside :)

[@DevOpsToolkit]

I think that you'll get a part of that picture in the next week's video.

[@DF-ss5ep]

Cdk8s is awesome

[@redfox6179]

Guys, gitlab or github for prespective which you want to choose?

[@DevOpsToolkit]

I prefer GitHub.com mostly because I'm focused on open source and almost all projects are in GitHub so I keep there both public and private repos.

[@redfox6179]

@@DevOpsToolkit thank, also second question, why in USA market are still more using jenkins, someone wrote that redflag it, but on "indeed " we can see a many vacancies with Jenkins requires.

[@DevOpsToolkit]

Jenkins was the de-facto standard for pipelines and was (and still is) used by many companies. People make massive investments in Jenkins and that means that moving to something else also requires massive investment. On top of that, no one really came up with a new approach to pipelines so all the solutions do, more or less, the same. As a result, there often are not sufficient reason to switch. You do not move from one thing to another if there are only marginal improvements.

[@ashimov1970]

Acorn and GitLab are missing

[@DevOpsToolkit]

Many others are missing :( I limited the scope to those that I used heavily in the past year or so. I prefer missing some than including those I cannot confirm for from personal hands-on experience. That should not reflect on the quality of those that are missing. Some are not there bacause they are bad choices while others are great but I cannot confirm that. I even excluded those that I used heavily in the past but I'm not up-to-date with their current status. In other words, there is a limit to how many tools I can use myself and I vouch for.

[@ashimov1970]

@@DevOpsToolkit I see... I suggest that you definitely give a consideration to Acorn. It's truely incredible

[@DevOpsToolkit]

Oh yeah. I used it but not yet long enough to judge it. I will spend more time with it soon.

[@ashimov1970]

@@DevOpsToolkit Can't wait for the Acorn overview 🙏🙂

[@DevOpsToolkit]

It's finally done: kzbin.info/www/bejne/q5uZeJSkZpWKn8k

[@lhxperimental]

AWS lightsail is not comparable to others in the list. It is just a VM provider like Linode, Digital Ocean or Hetzner

[@DevOpsToolkit]

You're right. That was a mistake on my part. I should have said "Lightsail Containers".

[@axedropgaming2544]

I really found this helpful, I also want to build career in devops can you please suggest me that from where I can start my career.

[@DevOpsToolkit]

Coding is the most important experience, followed by cloud (pick any of the big three), and kubernetes.

[@DevOpsToolkit]

What are your choices in 2023?

[@dreamcat4]

kata containers will be my one to watch in 2023. it rides on top of cloud hypeervisor (which is another newer modern project)

[@DevOpsToolkit]

Oh yeah. Kata Containers are great.

[@dreamcat4]

@@DevOpsToolkit hehe victor, i just realized these kata containers infrastructure could run on those new NVIDIA BlueField-3 dpu hardware. because it also supports the arm aarch64 architecture. pretty interesting for somebody out there with access to those new dpus. to make a better usage / utilization of them

[@cheebadigga4092]

I'll play around with Gateway API as soon as it graduates.

[@williamhardy1014]

I don’t understand: for HA why would you need replicas of an app? Because even if the app fails, kubernetes will handle it and auto restart it onto another agent node? Ensuring it is always available because it’s automatically restarted. Thanks

[@jemag]

A bit surprised by cdk8s. I feel like jsonnet + tanka is much simpler but people are simply less familiar with jsonnet.

[@DevOpsToolkit]

Familiarity with jsonnet and the support in IDEs are the main obstacles towards jsonnet adoption. I love how cdk8s imports Cards and converts them, in my case, into types go objects so code complete works like a charm not only for core kubernetes resources but for anything.

[@OlegKorsak]

nope, local PyCharm is much better than any of cloud envs

[@DevOpsToolkit]

PyCharm is an IDE. It won't give you a development environment (e.g., a Kubernetes with Pods and other resources that reflect your apps). I might be wrong though. I haven't used JetBrains in a while but, from what I understood, it does integrate with solutions like GitPod that do provide a full dev environment (e.g., apps running and synchronized).

[@kostiscodefresh]

In the same category as tilt/okteto/devspace you should also look at garden io and telepresence next year.

[@DevOpsToolkit]

Telepresence will be featured in a month or two in a new series dedicated to CNCF projects. Garden will have to wait a bit longer :(