As per my understanding B+ trees, Root and all the internal nodes are present in the main memory but the leaf nodes are present in HDD / SSD. Index is a data file so when we boot up MySQL it loads the index data file in the main memory and builds B+ tree and it resides in the main memory only the leaf nodes which contains the actual data( Page ) is present in the HDD/ SSD. As leaf contains the page ( which have multiple rows ) .
In LSM trees based indexing, data is added to the in memory mem tree which is implementing AVL tree or redblack tree structure which has very less amount of data in comparison to total data. All the other processes happen in the background. There is very minimal disk seek operation to find the place where the data is to be written, thus the individual write operation has a time complexity of around O(1) (number of nodes in AVL tree is much lesser than total number of data nodes in SS table segments), while the background compaction process adds the write complexity to O(N log N) time, where N is the total number of keys across SSTables. But compaction runs in background thread independent of write operation thread. Please correct me if I made a mistake here. This is as per what I have learned so far.
As per my understanding B+ trees, Root and all the internal nodes are present in the main memory but the leaf nodes are present in HDD / SSD. Index is a data file so when we boot up MySQL it loads the index data file in the main memory and builds B+ tree and it resides in the main memory only the leaf nodes which contains the actual data( Page ) is present in the HDD/ SSD. As leaf contains the page ( which have multiple rows ) .
FYI: Before watching this, please watch number 6 - LSM trees video
Thanks ,Informative video, I query: in case of secondary indexing . Does it store primary key or primary reference along with secondary key
How does breed maintains sequence of indexing?
Can you make a tutorial on how to implement in real time like for MERN or with some other stack?
Okay.. So.. Can i say that dbms is the same as systems design?
What is the video previous to this one, i am following the system design playlist. Seems confusing.
I dont understand how come LSM trees are sequential when they're storing data in a sorted fashion that makes it non sequential right?
sequential means the data is stored in contiguous fashion
In LSM trees based indexing, data is added to the in memory mem tree which is implementing AVL tree or redblack tree structure which has very less amount of data in comparison to total data. All the other processes happen in the background. There is very minimal disk seek operation to find the place where the data is to be written, thus the individual write operation has a time complexity of around O(1) (number of nodes in AVL tree is much lesser than total number of data nodes in SS table segments), while the background compaction process adds the write complexity to O(N log N) time, where N is the total number of keys across SSTables. But compaction runs in background thread independent of write operation thread.
Please correct me if I made a mistake here. This is as per what I have learned so far.
Thumbnail, i thought I clicked on dhruv rathee. 🤣
Bro do you still do CP?
No, I don't.
Happy married life
Hello brother can you kindly make a video how to crack HFT company off campus , how have you prepared brother??