Hello Milan, you content is just amazing, I like the fact that it is covering what you are trying to explain with all required examples and tools, so you can prove the results in practical way, keep going !!!
Hello Milan, thanks for all your videos and I have been following all your videos. I have a question in regards to consul service discovery as I have used it in the past. My question is is service discovery needed when you are using K8s which we can create deployment and specify replicas with load balancer? And we can also point to the service name rather than the IPaddress.
Great video. Next video idea: YARP load balancer + Consul with few service instances running P.S.: any plans on enhancing video quality to 1440p or 2160p ?
I think I'll write about YARP + service discovery first. 😁 Not planning on changing video quality for now. Which part is lacking? Code not clearly visible?
Interesting how everything was running on localhost yet Jaeger is showing interesting numbers there. Maybe because not configured for production? How will separate network calls if not on the same machine in distributed setup affect overall latency?
Typically the biggest cost there is network round trip time, which can be low if your services are in the same region. I run this via Docker Compose which makes a big perf impact.
Hey Milan awesome videos man thanks a bunch, I was trying out this example and I followed it to the dot, but some reason my services on the consul dashboard are giving this error:"TTL expired (last output before timeout follows): ttl", is there anything I am doing wrong ?
You are awesome. I have one question if I am using rabbitmq for communication between services using consumer. will it be applicable if there are like I have more than one instance of a service and ut consume HTTP call using rabbitmq
@@MilanJovanovicTech Sorry, let me rephrase my question again. I have 2 api (A and B) and in B I am using MassTransit.IConsumer along with rabbitmq. So that Api A can consume data from B using MassTransit.IRequestClient. Now I have two instances of Api B, So now when Api A will try to consume data from Api B then which instance of API B will give the response. In this can we use the same concept like consul
Want to master Clean Architecture? Go here: bit.ly/3PupkOJ
Want to unlock Modular Monoliths? Go here: bit.ly/3SXlzSt
Hello Milan, you content is just amazing, I like the fact that it is covering what you are trying to explain with all required examples and tools, so you can prove the results in practical way, keep going !!!
I appreciate that!
Thanks Milan, your last couple of videos have been quite timely.
Glad to hear that!
Thank you!
Sure thing
Amazing video, thank you!
Glad it helped!
Any videos about Dapr in the future?
Not for the moment
Thanks alot.
Glad to help
Is the physical address cached by consul for some time when you do the first lookup?
Consul does regular health checks to keep track if the downstream service is available
@@MilanJovanovicTech does a service using AddRoundRobinLoadBalancer() randomly select an instance for every request or is it fixed for the lifetime?
Hello Milan, thanks for all your videos and I have been following all your videos. I have a question in regards to consul service discovery as I have used it in the past. My question is is service discovery needed when you are using K8s which we can create deployment and specify replicas with load balancer? And we can also point to the service name rather than the IPaddress.
No, I don't think it's needed with K8s :)
Great video Milan. Can you also cover the .NET Aspire standalone dashboard for Open Telemtry.
Also in your sample app, the correct response for the POST request would be 201 created.
I already did: ua-cam.com/video/HrRrJ5wTtdk/v-deo.html
Great video.
Next video idea: YARP load balancer + Consul with few service instances running
P.S.: any plans on enhancing video quality to 1440p or 2160p ?
I think I'll write about YARP + service discovery first. 😁
Not planning on changing video quality for now. Which part is lacking? Code not clearly visible?
@@MilanJovanovicTech all is fine with video quality. But 1440 or 2160 is better 😜
Interesting how everything was running on localhost yet Jaeger is showing interesting numbers there.
Maybe because not configured for production? How will separate network calls if not on the same machine in distributed setup affect overall latency?
Typically the biggest cost there is network round trip time, which can be low if your services are in the same region. I run this via Docker Compose which makes a big perf impact.
Hey Milan awesome videos man thanks a bunch, I was trying out this example and I followed it to the dot, but some reason my services on the consul dashboard are giving this error:"TTL expired (last output before timeout follows): ttl", is there anything I am doing wrong ?
No idea, but grab the source code here and check: github.com/m-jovanovic/service-discovery-consul
Yarp can be used in combination with Consul in a Gateway service, I wonder?
Yes, but you'll have to implement it yourself. I'll probably try making a video about it.
@@MilanJovanovicTech Video would be handy
You are awesome. I have one question if I am using rabbitmq for communication between services using consumer. will it be applicable if there are like I have more than one instance of a service and ut consume HTTP call using rabbitmq
I don't get the second part of your question
@@MilanJovanovicTech Sorry, let me rephrase my question again. I have 2 api (A and B) and in B I am using MassTransit.IConsumer along with rabbitmq. So that Api A can consume data from B using MassTransit.IRequestClient. Now I have two instances of Api B, So now when Api A will try to consume data from Api B then which instance of API B will give the response. In this can we use the same concept like consul
This can be solved by messaging system …if no , then please explain
More latency, asynchronous communication. There are some downsides over plain HTTP communication.
I have never realized how similar this is to DNS
Yes, basically the same concept