At 12:55 you should push onto the heap before you pop, otherwise if the value you push is smaller than the value you are popping, your result will be incorrect.
really loved how you were suggesting solutions and evaluating them. Most solutions out there just touch a naive sql query solution and jump directly to a trie.
34:04 I don't see how the server has to be stateful. If I'm at "ap" and type in another "p", the server doesn't need to know that I was at "ap" previously, because now "app" is a new prefix query. So the server can be stateless. Is this not correct?
Question: @jordan @41:21, it is unclear to me how our queries are processed with the trie distributed across multiple servers. Lets say there are only two servers, one contains trie for words [a, b, aa,... abc] and other for [abcd,... zzzz]. In this case when I start typin and have typed "a", then I get connected to first server via websocket, and then when I type b and c, I continue to be connected and traverse the trie down to character 'c' on first server. Now, when I type 'd' what happens? Is it correct that: a) The first server (to which I am connected) will see that 'c' has no children, thus will break the websocket connection, therefore b) client will reconnect saying "gimme stuff for abcd", and load balancer directs me to second server, and c) the second server's trie has root node of "abc", so it goes down one step to 'd' and return the values? Is my understanding correct above?
I'd think that at the node at which the partition begins on system 1 it would say something like "no longer on this system" so that the client knows it needs to reach back out to the load balancer.
Great video Jordan. Had this question to ask: How does the design take care of server going down that was handling "app" (from word apple) partition. Trie data stored on that server would be lost so I believe data would be replicated across multiple nodes? Also, how would the switchover take place in this scenario? Zookeeper listening to all the active/replica nodes and switching over when it stops getting heartbeat? Is that correct understanding?
Yes, we'd want all nodes to have replicas. Zookeeper can listen to them and round robin requests to these nodes. In this case, all nodes are replicas as we don't write to any of them, they're all read only.
@ 21:08 short is 2bytes so its 16bits not 8 and hence we have like 65k terms ( i have to point this minor insignificant mistake or i cannot go to bed since I'm a internet police)
@@jordanhasnolife5163 I'm not sure I agree. Elasticsearch supports term and phrase suggestions as special use cases, and it gives users control over general relevance features. I work on a search team, and our design for this feature is centered around an Elasticsearch cluster w/ special typeahead indices, an ETL from BQ to that cluster, and a service to query the cluster. I don't know if our design is the industry standard, and it depends on exactly what you're trying to do, but I think this is definitely one of the ES use cases. (Typeahead isn't just about popularity either, there could be many different heuristics you need to use to rate which suggestions are the best. There may be machine learning models involved to help determine that as well.)
Hey Jordan, thanks for this video! I'm curious about how trie can be stored in different databases, could you share something related? Additionally, I think we can have cache in front of suggestion service, or probably what you mean by saying suggestion service already include that? Thanks!
I guess it can't really, hence our issue. You can be cheeky and use the tactic I used to store it in spark, however then you lose some of the nice time complexity As for Q2, the suggestion service is effectively already a cache - it's in memory and has cached the top suggestions for each search term
I think the main point is that HDFS is designed for processing large sequential data write like GB level, not good for small data write/read, thus message queue is better for keep event with small data size
The calculation of runtime at 13:40 is wrong. Runtime is O(log(n) + m) where is n is total words in dictionary, and m is total words matched by the query
I assume you're using this to account for the actual computation cost of finding the start and end of where we're reading. Good point. That being said, for the m piece I think it would be m * log(k), where k is the number of typeahead suggestions to return, as we need to formulate our heap.
A follow up video with a few more concepts would be great 1. We may not always need to reconnect a new web socket. That layer can be decoupled with suggestion service 2. How do we handle cases where updating the trie without impacting read performance- I’d guess something like update replica async and point traffic to it when ready? 3. What can be cached on client vs always query backend 4. Personalization / contextual results
@@jordanhasnolife5163 it's more good idea to create a video and just tagging the title as "Type Ahead Suggestion" and then not explaining any core concepts related to "Type Ahead Suggestion" and just going over few simple tree diagrams.
Another great video! Thanks for making it. I am a bit confused about the update path. 1. It looks like we are creating new trie from the logs (containing search term with freq in kafka) instead of updating the existing trie. Lets say we want to account for last few days of search, then to build the trie shouldn't we feed the copy of existing trie as well (along with recent search logs) to hdfs to calculate top suggestions for each prefix? 2. Instead of app server just getting data about top suggestion for each prefix from hdfs, is it possible for us to compute the trie as well offline and then load it in server? If yes, can you also please suggest tools to use for computing trie offline and loading from offline to server memory ?
1) HDFS already has the last few days of data available. It doesn't have to delete that just because we computed another trie from it. You wouldn't have to send the existing trie. 2) Considering that you can't really represent a trie in a text file like that, I'm not quite sure. I guess in theory, you could compute it on one server from the hdfs data, then serialize it to JSON or something, then send it out to all of the other servers. But then even, you're just building a trie from the JSON rather than the frequencies which frankly has a similar time complexity.
My girlfriend no joke asked if you were gonna steal me from her because of how much I talk about your channel. Keep it up
I'm Mr steal ya man
This is the best video I've seen explaining type ahead, thanks a lot for making great content!
At 12:55 you should push onto the heap before you pop, otherwise if the value you push is smaller than the value you are popping, your result will be incorrect.
This is true, nice catch
really loved how you were suggesting solutions and evaluating them. Most solutions out there just touch a naive sql query solution and jump directly to a trie.
Wow, a video where the capacity estimates actually matter. Really nice to see you compare these to memory amounts of client / servers.
I perform capacity estimates every weekend when figuring out how much late night food I should eat to not explode the next morning
34:04 I don't see how the server has to be stateful. If I'm at "ap" and type in another "p", the server doesn't need to know that I was at "ap" previously, because now "app" is a new prefix query. So the server can be stateless. Is this not correct?
Yes but then you have to traverse the trie again and ruin your time complexity
Question: @jordan @41:21, it is unclear to me how our queries are processed with the trie distributed across multiple servers. Lets say there are only two servers, one contains trie for words [a, b, aa,... abc] and other for [abcd,... zzzz]. In this case when I start typin and have typed "a", then I get connected to first server via websocket, and then when I type b and c, I continue to be connected and traverse the trie down to character 'c' on first server. Now, when I type 'd' what happens? Is it correct that: a) The first server (to which I am connected) will see that 'c' has no children, thus will break the websocket connection, therefore b) client will reconnect saying "gimme stuff for abcd", and load balancer directs me to second server, and c) the second server's trie has root node of "abc", so it goes down one step to 'd' and return the values?
Is my understanding correct above?
I'd think that at the node at which the partition begins on system 1 it would say something like "no longer on this system" so that the client knows it needs to reach back out to the load balancer.
Great vid as always! Would be cool to see how sentence suggestions are working, how words are connected to each other etc.
Pretty solid video. Really loved the optimisation you discussed.
Great video Jordan. Had this question to ask:
How does the design take care of server going down that was handling "app" (from word apple) partition. Trie data stored on that server would be lost so I believe data would be replicated across multiple nodes? Also, how would the switchover take place in this scenario? Zookeeper listening to all the active/replica nodes and switching over when it stops getting heartbeat? Is that correct understanding?
Yes, we'd want all nodes to have replicas. Zookeeper can listen to them and round robin requests to these nodes. In this case, all nodes are replicas as we don't write to any of them, they're all read only.
@ 21:08 short is 2bytes so its 16bits not 8 and hence we have like 65k terms ( i have to point this minor insignificant mistake or i cannot go to bed since I'm a internet police)
Lmao, well done, you've owned me
Why not something like Elasticsearch for prefix searching with the same range based partitioning?
It's going to be slower: that's on disk, and now I have to perform a binary search for my word rather than just traversing down a trie
@@jordanhasnolife5163 I'm not sure I agree. Elasticsearch supports term and phrase suggestions as special use cases, and it gives users control over general relevance features. I work on a search team, and our design for this feature is centered around an Elasticsearch cluster w/ special typeahead indices, an ETL from BQ to that cluster, and a service to query the cluster. I don't know if our design is the industry standard, and it depends on exactly what you're trying to do, but I think this is definitely one of the ES use cases. (Typeahead isn't just about popularity either, there could be many different heuristics you need to use to rate which suggestions are the best. There may be machine learning models involved to help determine that as well.)
Hey Jordan, thanks for this video! I'm curious about how trie can be stored in different databases, could you share something related? Additionally, I think we can have cache in front of suggestion service, or probably what you mean by saying suggestion service already include that? Thanks!
I guess it can't really, hence our issue. You can be cheeky and use the tactic I used to store it in spark, however then you lose some of the nice time complexity
As for Q2, the suggestion service is effectively already a cache - it's in memory and has cached the top suggestions for each search term
You should probably replace Flink with Spark Streaming since you already planning on using Spark downstream.
Yeah in reality I think that's reasonable, but for the sake of the systems design interview I like to be idealistic.
This is a great video, thanks for taking efforts to explain everything in such depth.
Curious why we need stream processing (Kafka -> Flink -> HDFS) to upload newly entered work to HDFS? Why cannot' we upload them to HDFS directly?
hdfs stores full files, not an individual string of text. We need to aggregate the queries first
@@jordanhasnolife5163 Is it better if we use spark streaming consumer instead of flink here? We can so batching using this and write a batch to HDFS
I think the main point is that HDFS is designed for processing large sequential data write like GB level, not good for small data write/read, thus message queue is better for keep event with small data size
Your jokes make the grind slightly less terrible :))
The calculation of runtime at 13:40 is wrong.
Runtime is O(log(n) + m) where is n is total words in dictionary, and m is total words matched by the query
I assume you're using this to account for the actual computation cost of finding the start and end of where we're reading. Good point.
That being said, for the m piece I think it would be m * log(k), where k is the number of typeahead suggestions to return, as we need to formulate our heap.
So many open questions left, need more clarity.
It's a good idea to leave a comment like this and then not ask any questions.
A follow up video with a few more concepts would be great
1. We may not always need to reconnect a new web socket. That layer can be decoupled with suggestion service
2. How do we handle cases where updating the trie without impacting read performance- I’d guess something like update replica async and point traffic to it when ready?
3. What can be cached on client vs always query backend
4. Personalization / contextual results
@@jordanhasnolife5163 it's more good idea to create a video and just tagging the title as "Type Ahead Suggestion" and then not explaining any core concepts related to "Type Ahead Suggestion" and just going over few simple tree diagrams.
Another great video! Thanks for making it.
I am a bit confused about the update path.
1. It looks like we are creating new trie from the logs (containing search term with freq in kafka) instead of updating the existing trie. Lets say we want to account for last few days of search, then to build the trie shouldn't we feed the copy of existing trie as well (along with recent search logs) to hdfs to calculate top suggestions for each prefix?
2. Instead of app server just getting data about top suggestion for each prefix from hdfs, is it possible for us to compute the trie as well offline and then load it in server? If yes, can you also please suggest tools to use for computing trie offline and loading from offline to server memory ?
1) HDFS already has the last few days of data available. It doesn't have to delete that just because we computed another trie from it. You wouldn't have to send the existing trie.
2) Considering that you can't really represent a trie in a text file like that, I'm not quite sure. I guess in theory, you could compute it on one server from the hdfs data, then serialize it to JSON or something, then send it out to all of the other servers. But then even, you're just building a trie from the JSON rather than the frequencies which frankly has a similar time complexity.
Top tier videos! Can you do Design a Parking Lot?
I already have!
Use public transport man, its getting costly to park cars these days!
@@alphabeta644 😂
Great video! Keep with the good job i really enjoyed it
What are your thoughts on using GraphDBs like Neo4j to store the trie?
I think that if we can avoid storing this guy on disk, we should! It's a pretty inefficient operation to jump from random spot to random spot on disk.
Merry Christmas 🎄🎁
Same to you!
Thanks for making the video. It was interesting and helpful.