Great Video!! I had one doubt, in this video, first we are uploading the complete video in blob storage and then through content processor we are chunking the video and doing multiple things with it. But the main concern is that how would we be able to upload let's say a 2 hour video directly to the s3. I feel the chunking should happen in the client only and client should send chunks to upload rather than client sending complete video to content upload service which is uploading the complete video first to blob storage and then chunking it and putting it back to blob storage again. Would love to know your thoughts on this
@adityamehta543 That is a wonderful question first of all. The answer to your question is at 12:14 where we talk about Resumable URL. Server sends a Resumable URL to client which it uses to upload the content. Now, even if it is 2 hour long video it doesn't trouble us. The upload gets paused and resumed based on the internet connection. But why aren't we chunking at the client at the very first place? Well, chunking on the client can be resource-heavy, especially on low-end devices or poor networks. By chunking at the server, we simplify the client’s workload and get better control over how we process and store the video. That said, client-side chunking can be useful in some cases, especially for large uploads. It’s a trade-off depending on the use case.
@@sweetcodey Thank you so much for answering the question! This truly helps and also the video explaination is amazing. I have few more concerns also, if you can take it up-: 1) Can you also talk little bit about search functionality here, if we have to include search service, how would it interact with elastic search and will elastic search get the data from video db and then index on it and return ? Just how it will look like if you can tell over text only. 3) Uploading a video is a functional requirement , but notification is also functional requirement? Thank you so much and sorry for bombarding the comment section with questions. The video is very detailed and I am learning a lot from it.
@adityamehta543 Thanks for the appreciation! 1) Regarding the search functionality, it might be more than what’s needed here. In an interview setting, we recommend focusing on core features like upload and stream. Search is an extensive topic on its own, and we’ll dive deeper into it in future UA-cam videos and courses at sweetcodey.com. 2) As for notifications, they can be a nice touch if the interviewer brings it up, though it's not essential. It wasn’t part of the original functional requirements, but we think it’s a valuable addition if time permits.
Great Video!! I had one doubt, in this video, first we are uploading the complete video in blob storage and then through content processor we are chunking the video and doing multiple things with it. But the main concern is that how would we be able to upload let's say a 2 hour video directly to the s3. I feel the chunking should happen in the client only and client should send chunks to upload rather than client sending complete video to content upload service which is uploading the complete video first to blob storage and then chunking it and putting it back to blob storage again. Would love to know your thoughts on this
@adityamehta543 That is a wonderful question first of all. The answer to your question is at 12:14 where we talk about Resumable URL. Server sends a Resumable URL to client which it uses to upload the content. Now, even if it is 2 hour long video it doesn't trouble us. The upload gets paused and resumed based on the internet connection.
But why aren't we chunking at the client at the very first place? Well, chunking on the client can be resource-heavy, especially on low-end devices or poor networks. By chunking at the server, we simplify the client’s workload and get better control over how we process and store the video. That said, client-side chunking can be useful in some cases, especially for large uploads. It’s a trade-off depending on the use case.
@@sweetcodey Thank you so much for answering the question! This truly helps and also the video explaination is amazing. I have few more concerns also, if you can take it up-:
1) Can you also talk little bit about search functionality here, if we have to include search service, how would it interact with elastic search and will elastic search get the data from video db and then index on it and return ? Just how it will look like if you can tell over text only.
3) Uploading a video is a functional requirement , but notification is also functional requirement?
Thank you so much and sorry for bombarding the comment section with questions. The video is very detailed and I am learning a lot from it.
@adityamehta543 Thanks for the appreciation!
1) Regarding the search functionality, it might be more than what’s needed here. In an interview setting, we recommend focusing on core features like upload and stream. Search is an extensive topic on its own, and we’ll dive deeper into it in future UA-cam videos and courses at sweetcodey.com.
2) As for notifications, they can be a nice touch if the interviewer brings it up, though it's not essential. It wasn’t part of the original functional requirements, but we think it’s a valuable addition if time permits.