Designing Uber's highly available Emergency SOS Service

  Рет қаралды 9,758

Arpit Bhayani

Arpit Bhayani

Күн бұрын

Пікірлер: 12
@manmohanmundhraa3087
@manmohanmundhraa3087 Жыл бұрын
Solution you mentioned is using server side logic. From user mobile, uber app, can sent email, call to 911, message to emergency contacts. This will be quickest. From server side some issue may happen.
@yolo420swag69haboub
@yolo420swag69haboub 5 ай бұрын
Would be nice if you shared the source of where you got your info on the architecture
@koteshwarraomaripudi1080
@koteshwarraomaripudi1080 2 жыл бұрын
Great design, especially having a fallback to make sync calls in case of Kafka goes down 👌. It would be interesting to know about their deployment strategy as they can't afford any downtime (I'm assuming blue-green would be a good fit here). Thank you @Arpit Bhai for bringing such amazing content for us. I don't think there is any other channel that posts such amazing system design videos as you do.
@abhis3kh
@abhis3kh 2 жыл бұрын
Hi Arpit, Thanks for efforts :). Really appreciate it. Have some Questions: 1. Main issue in my mind: What if the network is not good, is there any way we could draft a local SMS and send to SOS contacts with last cached information from the uber app itself? (Wanted to know your views on it.) 2. Should we fetch the data from Server side (like Driver/trip details etc) or we should use the cache as it also has the same data if we connected through the internet continuously? (Taking the data from server will take some time) 3. How the RAPID-SOS works? How it captures the nearest local police location? 4. What does this 3PT means? 5. As we are broadcasting the data continuously, how to know that emergency is gone & we will stop sending the alerts? & What you use drawing these diagrams?
@AsliEngineering
@AsliEngineering 2 жыл бұрын
1. Yes 2. Both should be done with server being the source of truth 3. RapidSOS is a company and it is their headache on how they do it. Fir uber it is a blackbox.
@AsliEngineering
@AsliEngineering 2 жыл бұрын
4. 3rd party 5. Someone from uber team is coordinating throughout
@AsliEngineering
@AsliEngineering 2 жыл бұрын
GoodNotes
@abhis3kh
@abhis3kh 2 жыл бұрын
@@AsliEngineering THank you :). If possible, could you make a video on RAPID SOS architecture :)
@jatinthakwani5370
@jatinthakwani5370 Жыл бұрын
Hi Arpit, Great content. You mentioned that uber would be continuously storing data of our location, so the database operations need to be extremely fast because there would be a lot of load on the database when the SOS button is clicked, same goes for read speed because we would need the latest data to render it on the uber support team. If you would be designing such a system, which database you would choose to maintain high read and write speeds?
@AsliEngineering
@AsliEngineering Жыл бұрын
There are bunch of videos on my channel for talking about this. Go through them.
@rjarora
@rjarora Жыл бұрын
One of your rare videos where there was not much new to learn technically
@ImRachitB
@ImRachitB 11 ай бұрын
You must be a genius to know-it-all.
3. Observer Design Pattern Explanation, Walmart Design Interview Question, 2022 | LLD System Design
34:34
From Small To Giant 0%🍫 VS 100%🍫 #katebrush #shorts #gummy
00:19
Lazy days…
00:24
Anwar Jibawi
Рет қаралды 8 МЛН
How to Fight a Gross Man 😡
00:19
Alan Chikin Chow
Рет қаралды 21 МЛН
16. System Design - Distributed Messaging Queue | Design Messaging Queue like Kafka, RabbitMQ
45:13
How Flipkart made their type ahead search hyper personalized
19:00
Arpit Bhayani
Рет қаралды 15 М.
Razorpay's Journey to Microservices w/ Arjun | Ep 1
49:48
Arpit Bhayani
Рет қаралды 56 М.
Why, where, and when should we throttle or rate limit?
19:05
Arpit Bhayani
Рет қаралды 15 М.
How Razorpay scaled their notification system
17:32
Arpit Bhayani
Рет қаралды 22 М.
How to learn better and faster as a software engineer?
14:57
Arpit Bhayani
Рет қаралды 40 М.
How @zomato improved its search by identifying intent using NLP
26:45
How Gojek masks and keeps users' phone numbers secure at scale?
16:21