thanks for the informative video. One correction though, REST is not a protocol like HTTP, it is an architectural approach/a set of rules to develop API. In fact HTTP is the underlying protocol for RESTFull API.
There is a fourth entry point, which is the Lambda end point (which can be exposed externally). I'm doing small stuff, but that's what I use. I wonder how that fits in - maybe just for 'small stuff'?
Hello - thank you for our video! I have an architecture question. If I have a public API and private API on the same server exposed by an internal ALB - the public API is exposed by a public API Gateway (resides outside the VPC), where as the private api is exposed by a Private API Gateway which is accessible through a VPN. Both of the API Gateways point to the internal ALB. Does this set up make sense to your or is there a better way to have a private API and publci API accessible to end users on the same server?
1st of all thank you for wonderfully organized and immediately useful contents! 2nd - there is a not publicly-advertised (but not secret) capability to increase gateway timeout beyond 30 seconds. This quota increase is ONLY available to Premium (Enterprise) Support customers. Customer does have to agree to lower RPS when API Gateway timeout is increased the formula is pretty simple. So if you want one minute timeout your RPS will be halved. (former Senior TAM here).
@@ankitadixit5875 there are similar roles in other clouds. There is a pathway to become s Solutions Architect, or maybe Account Manager (more sales-focused).
Thanks for the video. I've been using REST API gateway for a few years now, but the rest of the details about http and websockets were insightful. One remark, I think that there might be few more limitations to API Gateway. I think that there is a different limit for data response maybe? (5 mb?)
Great video. Question for you: if I have an API Gateway connected to a REST API lambda, and I wanted to use a custom rate limiter lambda to check req count, would it be possible to have the gateway hit the rate limiter first to verify if the user is allowed to send the request, then hit the API? Kind of like an authorizer
You can make the API Gateway hit a REST API Lambda that would be used as a middleware count checker first before routing the request to a second one that would handle user request. It's to optimize AWS profits, trust me, i know how to scale costs
@@tzvimashiach7728 Bro I'm so sorry that was meant to be a joke =/ My main objective as a software engineer is to scale costs, not performance ! So I do make Lambdas call some Lambdas, this way, they themselves can call other Lambdas etc.. Anyway, there is a AWS documentation about API Gateway rate limiting, check it out ! Good luck have fun my fellow human
I thought this video was about AWS API gateway, not HTTP, REST, or Websockets. If you want to talk about those technologies, you should create a video about them.
AWS API Gateway is used to create custom HTTP, REST, and WebSockets API’s without the need to setup a lot of the authentication and endpoints and stuff.
FYI: the underlying protocol for REST is HTTP. So, there is no protocol called "REST".
youtube didn't just place this gem in my recommendations. subscribed
thanks for the informative video. One correction though, REST is not a protocol like HTTP, it is an architectural approach/a set of rules to develop API. In fact HTTP is the underlying protocol for RESTFull API.
hi! AMazing video!!
where can we find how to set it up ( like a tutorial )
Great video
Awesome tutorial on API Gateway Fundamental tutorial! Thanks for sharing. 👏👏👏👏
There is a fourth entry point, which is the Lambda end point (which can be exposed externally). I'm doing small stuff, but that's what I use. I wonder how that fits in - maybe just for 'small stuff'?
Hello - thank you for our video!
I have an architecture question. If I have a public API and private API on the same server exposed by an internal ALB - the public API is exposed by a public API Gateway (resides outside the VPC), where as the private api is exposed by a Private API Gateway which is accessible through a VPN. Both of the API Gateways point to the internal ALB. Does this set up make sense to your or is there a better way to have a private API and publci API accessible to end users on the same server?
1st of all thank you for wonderfully organized and immediately useful contents!
2nd - there is a not publicly-advertised (but not secret) capability to increase gateway timeout beyond 30 seconds. This quota increase is ONLY available to Premium (Enterprise) Support customers. Customer does have to agree to lower RPS when API Gateway timeout is increased the formula is pretty simple. So if you want one minute timeout your RPS will be halved. (former Senior TAM here).
are there any scopes for AWS TAM outside of AWS, I meant at professional level?thanks
@@ankitadixit5875 there are similar roles in other clouds. There is a pathway to become s Solutions Architect, or maybe Account Manager (more sales-focused).
Excellent overview. Thank you.
Thanks for the video. I've been using REST API gateway for a few years now, but the rest of the details about http and websockets were insightful.
One remark, I think that there might be few more limitations to API Gateway. I think that there is a different limit for data response maybe? (5 mb?)
There is also a 5k concurrent burst calls and 10k calls/sec
Great overview, thank you
25:20 it's per million :D
Very good content
good job
Thanks
Great video.
Question for you: if I have an API Gateway connected to a REST API lambda, and I wanted to use a custom rate limiter lambda to check req count, would it be possible to have the gateway hit the rate limiter first to verify if the user is allowed to send the request, then hit the API? Kind of like an authorizer
You can make the API Gateway hit a REST API Lambda that would be used as a middleware count checker first before routing the request to a second one that would handle user request. It's to optimize AWS profits, trust me, i know how to scale costs
@@randomdev4653 Thanks! Would the API response go back through the API/rate-limiter lambda? Why not just put it all into one lambda in that case?
@@tzvimashiach7728 Bro I'm so sorry that was meant to be a joke =/ My main objective as a software engineer is to scale costs, not performance ! So I do make Lambdas call some Lambdas, this way, they themselves can call other Lambdas etc.. Anyway, there is a AWS documentation about API Gateway rate limiting, check it out ! Good luck have fun my fellow human
I thought this video was about AWS API gateway, not HTTP, REST, or Websockets. If you want to talk about those technologies, you should create a video about them.
You should rethink that statement cuz it' makes no sense
?
AWS API Gateway is used to create custom HTTP, REST, and WebSockets API’s without the need to setup a lot of the authentication and endpoints and stuff.
please stop using such stupid services, just have your own api and do your thing. it's gonna be much faster believe me
Newbie question but is it worth the added maintenance and what’s the typical cost difference for similar request rates?