Communication patterns are very clearly explained in the video but I am still confused on how reactive model fits in this picture? I mean, when we are using http rest, if we use reactive programming model, will the communication between services not become async? If so, will we still need a queue structure to make the services decoupled or is just using reactive model enough for the services to be decoupled and communicate asynchronously? Thanks again.
While using JWT we have secured microservices from external clients, but few questions arises: 1) How will a microservice communicate with another microservice (will they need a JWT, how to create then?) or to keep such communication unsecured? 2) How to protect a microservice from direct access by clients? How to make sure gateway is the only way to reach a microservice?
In most cases, microservice-to-microservice communication does not require JWT. Once the user ID and other relevant information are extracted from the JWT by the first microservice, they can be passed to the subsequent microservices as needed. To secure microservice-to-microservice communication within a Kubernetes cluster, you can employ trusted solutions like service meshes, which help secure communication using mutual TLS (mTLS). Service meshes like Istio also provide features such as service discovery. By default, in a Kubernetes cluster, all microservices are defined using ClusterIP (best practice), which means they are not exposed to external communication. However, if you need to expose a microservice externally, you can use an Ingress resource with a Layer 7 load balancer. API Gateways like Kong and Traefik can manage the load balancer and handle the routing of external requests to the appropriate microservices. It's important to note that my answer is primarily in the context of using a Kubernetes cluster. I don't have extensive experience with microservice communication in other environments.
I am facing a real world problem: I have a PDF generation utility who time consumption directly propotional to the number of pages generation. This is implemented in microservice architecture. Would async approach solve the problem?
Absolutely! A request response implemented with async approach can help it. Send a link to download the pdf or a activate the button to download the pdf after the generation is over.
Hey, the video was a great refresher on communication types b/w services. As a next step (maybe discussed later in series) can you explain about mediator and adapter layers in micro-services? Also, thanks for taking out time and making such awesome videos!!
Great video! I still don't understand what if a microservice needs data from another. For example, the endpoint in Microservice 1 needs data from Microservice 2. How will Microservice 1 get the data needed in event-driven architecture to send back the response for the user?
Great video!! Can you please explain why shared DB approach is not good one when more than 2 services write and read? If you can provide more detail on the drawbacks?
Event based approach some hiw aounds like drop msg and forget kind of concept. What will happen when some of the downstream service fails . How to handle such scenario. Thanks in advance
These videos are so rich with years of experience of genuine microservice development
One small video with so much content in simple language with better visualizations.
Hats Off Mam
Getting addicted to your videos.
Sounds interesting definitely looking forward to get going with your micro -services videos.
Thank you for insightful kickstart.
Thanks Yogita, a dedicated series on Event Driven & Microserice Architecture is really good for us.
Another amazing tutorial Yogita. Your tutorials have been really helpful. You're the best😊
Hi Yogita thanks for dedicated videos on microservices, although an arranged playlist specifically for microservices would really help :) thanks
+1
the video is 8 min long and it has dense info in it, you are very helpful thanks.
Since long time i am searching these type of video, Please make the video serieas of microservices
Thanks for this video, will be helpful if we get to see more on event driven architecture
Communication patterns are very clearly explained in the video but I am still confused on how reactive model fits in this picture? I mean, when we are using http rest, if we use reactive programming model, will the communication between services not become async? If so, will we still need a queue structure to make the services decoupled or is just using reactive model enough for the services to be decoupled and communicate asynchronously? Thanks again.
got a good understanding , nice video!!!
Hi Yogita can you please make a video playlist on microservices best practices
While using JWT we have secured microservices from external clients, but few questions arises:
1) How will a microservice communicate with another microservice (will they need a JWT, how to create then?) or to keep such communication unsecured?
2) How to protect a microservice from direct access by clients? How to make sure gateway is the only way to reach a microservice?
In most cases, microservice-to-microservice communication does not require JWT. Once the user ID and other relevant information are extracted from the JWT by the first microservice, they can be passed to the subsequent microservices as needed. To secure microservice-to-microservice communication within a Kubernetes cluster, you can employ trusted solutions like service meshes, which help secure communication using mutual TLS (mTLS). Service meshes like Istio also provide features such as service discovery.
By default, in a Kubernetes cluster, all microservices are defined using ClusterIP (best practice), which means they are not exposed to external communication. However, if you need to expose a microservice externally, you can use an Ingress resource with a Layer 7 load balancer. API Gateways like Kong and Traefik can manage the load balancer and handle the routing of external requests to the appropriate microservices.
It's important to note that my answer is primarily in the context of using a Kubernetes cluster. I don't have extensive experience with microservice communication in other environments.
nice explanation...
Could you make a video on distributed logging mechanism between micro services
I am facing a real world problem:
I have a PDF generation utility who time consumption directly propotional to the number of pages generation.
This is implemented in microservice architecture.
Would async approach solve the problem?
Absolutely! A request response implemented with async approach can help it. Send a link to download the pdf or a activate the button to download the pdf after the generation is over.
@@sudocode I got this idea when I was watching your video. I will try this. Thanks for sharing.
@Jagruti Tiwari, you can use hangfire here as it is best approach for background processing jobs
Hey, the video was a great refresher on communication types b/w services. As a next step (maybe discussed later in series) can you explain about mediator and adapter layers in micro-services?
Also, thanks for taking out time and making such awesome videos!!
Great video! I still don't understand what if a microservice needs data from another. For example, the endpoint in Microservice 1 needs data from Microservice 2. How will Microservice 1 get the data needed in event-driven architecture to send back the response for the user?
love it
Great video!!
Can you please explain why shared DB approach is not good one when more than 2 services write and read? If you can provide more detail on the drawbacks?
Microservices should be independent but sharing database doesn't make it though
Amazing video🙌 Could u plz give a little more deep insight on which approach should be used when? It'll be very helpful.
Good tutorial
helpful thank you so much
Thank you very much
can you please shed some light on Soa architecture.. thanks in advance😄
Event based approach some hiw aounds like drop msg and forget kind of concept. What will happen when some of the downstream service fails . How to handle such scenario. Thanks in advance
Liked the video.Please start a tutorial on guitar. Maybe i will finally learn that😢
What if the event never get's delivered to the downstream services? What if there were packet loss?
Covered in next video
Any chances that we're going implement these things 😀
what about graphQL call ?
Can you make one video on redis
In last approach service don't communicate. Then how notification service know that there is a order
event-driven architecture
Madam you are simply super talented and beautiful girl