14 күн бұрын
4 ай бұрын
6 ай бұрын
6 ай бұрын
7 ай бұрын
7 ай бұрын
8 ай бұрын
Пікірлер
@dharanyuvi6951 3 күн бұрын
What if GC got AI powered in future...can it change the game for java?
@TheGeekNarrator 3 күн бұрын
GC has already significantly gotten better. Look at the new GCs available that are also optimised for massive heaps. About AI I am not sure how that would work and what will be the overhead of it. But only time will tell. Currently there are few applications that can have problems even with minimal GC pauses like stream processing, for those I think a non-GC language might be a better choice. But again its all about trade offs and there is no single answer.
@thatguyadarsh 22 күн бұрын
Very detailed and to the point discussion. I am in awe as this is just what I needed. Very grateful for this discussion. Thanks for sharing!!
@TheGeekNarrator 22 күн бұрын
Thanks for watching
@VahidOnTheMove 23 күн бұрын
How about to edit/delete a record?
@arun553 25 күн бұрын
absolute heat, thanks! I'm a big fan of the highlighted subtitles fwiw
@RandhirKrSingh-x9l 26 күн бұрын
Thank you for creating this video Kaivalya! I learned so much about DynamoDB in just 90 minutes from you and Alex. Such valuable insights!
@arun553 29 күн бұрын
solid, ty! The subtitles came in handy, I was pretty much relying on those when playing at faster playback speeds!
@hariharapamarjane2215 Ай бұрын
How many parts will be there?
@TheGeekNarrator Ай бұрын
At least 5. But will keep adding to this series. There is just so much to talk about 😀
@Alpheus2 Ай бұрын
Cool format. Pleasantly surprising how much info you can convey in a fun way in under 20mins.
@PraveenKumarBN Ай бұрын
This is amazing... Keep coming up with more content, especially on Graph Databases
@hanamufidah5809 Ай бұрын
Thank you both for the vid. It's a great discussion. 30:02, Instagram allows user to edit the post dynamic content (description, caption, location, tags, etc)
@VaibhavPatil-rx7pc Ай бұрын
Awesome explaination 🎉, thanks
@abhishekshrivastav7941 Ай бұрын
He has not spent more time in industry where extremely busy IC schedule open office hours so that their work doesn't get impacted by each and everyone disturbing them.
@souravdhar47 Ай бұрын
❤ this content is just awesome
@mkres Ай бұрын
am I the only one questioning how packing 10.000 business transactions into a single DB transaction works consistency wise? :)
@DurgeshGautam-d2b Ай бұрын
Bb
@harshavardhanreddykhvr 2 ай бұрын
How is Hdfs different from the above mentioned approach?
@TheGeekNarrator 2 ай бұрын
HDFS is a file system while we talked about object storage. File system is hierarchical while object storage is flat, which allows for infinite storage and scalability. File system on the other hand has limitations in that regards. File system might be better in terms of speed and performance but again depends on the total size etc. so yes the approach we discussed is purely on top of cloud object storage.
@KavirajKanagaraj 2 ай бұрын
What is the rust library Simon talked about building flexible storage formats? I heard something like "archive". But failing to find the right one.
@TheGeekNarrator 2 ай бұрын
Good question github.com/rkyv/rkyv
@simon_eskildsen Ай бұрын
rkyv
@manish-mk 2 ай бұрын
These podcasts are goldmine! Also I watched your only livestream today while I was in metro. Got many insights. You should do more livestreams.
@TheGeekNarrator 2 ай бұрын
Thanks 🙏🏻 I will make a note for livestreams.
@Md_sadiq_Md 2 ай бұрын
Pushing the algorithm ❤
@himanshupandey9303 2 ай бұрын
Thanks for updating the video. Great content. Keep it up.
@TheGeekNarrator 2 ай бұрын
Your feedback helped me. thanks :)
@AmeliaArdath 2 ай бұрын
Hi! So, I just integrated DuckDB into my little open source distributed trace viewer and came here to learn more about its internals, but I have to stop the talk for a second and thank you profusely for having clear, amazing subtitles, right in the video. As a programmer with profound hearing loss, it's usually a struggle for me to get through technical videos with the jumpy auto-generated captions KZbin provides, and it so rare that I get to just sit back and enjoy a talk like a hearing person would. Thank you for providing this experience!
@TheGeekNarrator 2 ай бұрын
This comment has made my day. Thanks a lot for sharing that. Unfortunately I got a lot of comments asking me to remove those subtitles as they were distracting so I had to remove them from other videos. Now I am motivated to find a way to make it work for both use cases.
@HimanshuKanodia 2 ай бұрын
👍
@HimanshuKanodia 2 ай бұрын
👍
@HimanshuKanodia 2 ай бұрын
👍
@HimanshuKanodia 2 ай бұрын
I don't build product directly, but in kind of integration work, but honestly even if I try to think from user perspective, I don't see people around me, in teams, willing to work like that or let others do. 🤣🤣
@HimanshuKanodia 2 ай бұрын
Nice podcast. Today only I started listening your podcast.
@VipulVaibhaw 2 ай бұрын
This was awesome!
@TheGeekNarrator 2 ай бұрын
Thanks a lot Vipul 🙏🏻
@bbowjazz 2 ай бұрын
Heikki rocks!
@aneksingh4496 2 ай бұрын
Your podcasts and the way you dig down to detailed level is revolutionary...please keep making such insightful videos ...this has enhanced my knowledge many folds .. let's catch up if you want
@anthonya880 2 ай бұрын
Please remove members only videos. It doesn't make sense yet for your channel.
@TheGeekNarrator 2 ай бұрын
Thanks for the feedback. I am happy you added “yet”, which tells you believe 😀🙏🏻
@anthonya880 Ай бұрын
@@TheGeekNarrator Yes your channel is different. It has lot of potential.
@RokijulIslam-sk4dt 2 ай бұрын
😅o
@D9ID9I 3 ай бұрын
The proper way is to rewrite using native languages like C++ or Rust. Use of java again is a plain joke.
@kaimingwan526 2 ай бұрын
Language has never been an issue with Kafka. Embracing the Kafka ecosystem and being fully compatible with Kafka is an important trade-off.
@D9ID9I 2 ай бұрын
@@kaimingwan526 Efficiency is the issue. It translates to number of resources you have to pay for monthly. That's why Redpanda exists.
@costathoughts 3 ай бұрын
Thank you Kaivaly for the video, it was really well explained and the animations are amazing. I would like to point just one thing the title "Reinventing Kafka the right way" I think is not precise once the part that AutoMQ are improving is the Storage Engine system and not the Message Middleware Broker (if I understood correctly based on this video). It was more an opportunity to enhance an isolated component from Kafka for cloud storage instead a full rewrite. An example of "Reinventing the right way" in my point of view would be the ScyllaDB which has backwards compatibility with Cassandra although it was rewritten using C++. I wish the best brow keep going!!!! EDIT 1: And they created a platform over na existent Kafka concept
@TheGeekNarrator 3 ай бұрын
Thanks for watching and the feedback. I appreciate it. I think you made a fair point, I guess re-engineered would have been a more precise word. I will change that.
@ShortGiant1 3 ай бұрын
Great video, learnt a lot!
@nosh3019 3 ай бұрын
Thanks for this interview. Great as always! One question: so if the WAL writes are batched and periodically flushed to storage, there in a window in which if the process crashes, the writes will be lost? I.e the client response that write is complete is not durable.
@TheGeekNarrator 3 ай бұрын
Thanks for watching. Yes, there are ways to configure the behaviour. This might help you slatedb.io/docs/faq/#what-happens-if-the-process-goes-down-before-slatedb-flushes-data-to-object-storage
@ShortGiant1 3 ай бұрын
Yeah that’s the “linger” stuff they spoke about. The client won’t get an ack about the write being successful until the flush happens. Therefore it would be as if the write request never happened (in practice the client would retry)
@harshrai456 3 ай бұрын
The storage architecture reminds me of AWS RDS (Aurora) Architecture with the control plane managing a muli tenant storage fleet and having write quoram for strong consistency. Great discussion, loved it.
@Jasin-p6w 3 ай бұрын
Wonderful video
@jmitesh01 3 ай бұрын
Information dense conversation! And isn't it expected that size is sufficiently large then it would increase number of put calls to object storage. Why don't you do buffering based on time plus size based criteria?
@TheGeekNarrator 3 ай бұрын
Good question. I think there are plans but haven’t really finalised yet. But maybe in the future a hybrid setting will be available. IIRC I asked this question to Chris.
@thrawn01 3 ай бұрын
Thank you for this, looking forward to assisting with the golang port! This was a great overview of the project, and a great way to get acquainted with some of the design decisions.
@TheGeekNarrator 3 ай бұрын
Thanks for watching 🙏🏻
@plargyle 3 ай бұрын
I love how you simplify complex topics in the crypto space.
@MarkHarrison-g7r 3 ай бұрын
Like all duckdb users, I really enjoy hearing Mark Raasveldt talk. You did a great job of guiding the conversation into intesting and informative areas!
@TheGeekNarrator 3 ай бұрын
Thanks a lot watching and sharing your thoughts 🙏🏻😀
@arkadutta744 4 ай бұрын
This is super cool . Probably the only youtube video [> 1hr ] , which I watched in whole. I took a whole week to watch it ... made sure I understood every idea presented and discussed. Thank you so much. I have a question => The partition placement , is DDB still using consistent hashing , for deciding which partition goes to which node[physical server] ... or there is some other algorithm used now. The original Dynamo paper mentioned consistent hashing [I read half the paper] ... Precisely => 1. How is it decided which key goes to which partition ? 2. How is it decided which partition goes to which node[physical server] ? Consistent hashing ... works in a bit different way ... all of the people watching this video knows how . For example in normal consistent hashing I guess it is hard to keep a limit on partition size ... I have read some part of the original consistent hashing paper ... in my quest to understand that paper ... somehow I reached your video .. and I thank myself for that.
@imranfp 4 ай бұрын
Well presented and explained well. The best part when Matthias use pointer to exactly mention what part in presentation is under discussion. Thanks 👏