Really like your content! For suggestion in python's ecosystem django is the most dominant backend framework, I know It will get beaten by node, java and golang in performance but still we would love to see how slow it is compared to other languages. Tip: In django most of the people use DRF (Django Rest Framwork) for building REST APIs so combining django with DRF makes sense when doing benchmark.
Hi, I love your videos! Would you consider making a video about GitOps strategies and best practices in a closed, bare metal environment (without using a cloud solution like GitHub)? Specifically, how do you host your infrastructure Git repo? Do you create a separate non-GitOps cluster to deploy HA-Git and CI/CD? Do you deploy Git to a plain VM? Or do you go through the "chicken and egg problem" and deploy Git and CI/CD into the same cluster that you manage with GitOps tools like Flux or Argo?
Hello Anton, thanks for the video, great as always. If you have time and possibility can you please make videos a little bit longer with deeper explanations for such noobs like me?)) Do you plan any videos on such topics as jenkins, ansible? Anyway thanks, you are doing great job.
Thanks Anton, amazing as always. I can see you are using Grfana stacks for these statistics and reports, what about Elastic stack do you think Grafana stack is better? Could you please compare both solutions: Elastic stack with metric beat and Opentelemetry Elasticsearch exporter and Grafana with tempo?
Grafana was one of the first to support Prometheus as a data source. For ELK, you need to use Elastic to store metrics, and the pipeline would be somewhat different and more complex, but, I will definitely create a comparison video at some point.
@@AntonPutra Yep, completely agree. It’s kind a common use case to use Elastic or Opensearch in many companies, so I was wondering why. Thanks for the reply, awesome as always!
@@valerikehayov2393 initially it was only for logs, but ELK trying to create universal solution for metrics/traces/logs but it's expensive. Try out loki, it indexes only labels and much cheaper to host
Thank you for an amazing content. Could you one day make a tutorial about how to set up multiple node k8s cluster locally: ESXi or lxd if possible. It would help to learn k8s without AWS stuff so much.
@@AntonPutra That is a good starting point to begin the journey but these nodes would be docker containers on the same host/VM. I consider building a k8s cluster manually on different nodes (physical hosts/VMs) to play with it.
@@AntonPutra Thank you! You used ESXi in this video and since I have a dedicated ESXi host where I build virtual network topologies and practice my programming skills I thought it would be great to have a small k8s lab there too to (you know) some kind of versatile IT experience =)
Do the performance benchmark for Quart please! It's the async version for Flask so you can use your Flask code and modify it a lil bit. Plenty of Flask extensions are working in Quart too! Cheers!
@@AntonPutra sure, add constant like 50-150ms for both languages and the difference would almost disappear. But CPU/MEM difference would not, for sure)
I wonder how FastAPI would perform, compared to Flask. I am assuming that since it supports asynchronous processing of requests, contrary to Flask (ASGI vs WSGI), it should overall reduce latency.